> ## Content Index > Fetch the complete content index at: https://www.ceo.com/llms.txt > Use this file to discover other available public pages before exploring further. # Endor Labs Founder & CEO Varun Badhwar: Why AI makes cybersecurity more important than ever - URL: https://www.ceo.com/episodes/endor-labs-founder-ceo-varun-badhwar-why-ai-makes-cybersecurity-more-important-than-ever/ - Published: 2026-06-10T10:00:27.000Z - Updated: 2026-08-22T21:54:42.000Z - Description: The Endor Labs CEO on securing code nobody wrote, trusting maintainers nobody has met, and rewriting a trillion lines. - Author: Clint Betts # The end of hidden vulnerabilities For decades, companies made a quiet bet: ship the software now, fix the security later, and hope nobody finds the flaws. Varun Badhwar says the bet is over. "The reality today is that advanced AI systems will find virtually every security issue hidden inside software," the Endor Labs founder and CEO said on the CEO.com Podcast. Anthropic reportedly has a model, Mythos, that it won't release because it finds vulnerabilities too well. Badhwar is unmoved by the restraint. "Whether Anthropic releases Mythos or not doesn't really matter. The capability already exists." OpenAI has it. Google has it. So do open-source labs in China. Defenders have to be right every time. Attackers only have to be right once. Badhwar has spent 20 years building security companies at moments when a technology shift made them necessary. When enterprises moved to SaaS around 2010, he founded CipherCloud. When they moved to the cloud, he founded RedLock, which [Palo Alto Networks bought for approximately $173 million](https://investors.paloaltonetworks.com/news-releases/news-release-details/palo-alto-networks-announces-intent-acquire-redlock?ref=ceo.com) in 2018, three years after its founding. He stayed on to build Prisma Cloud into one of the company's flagship businesses. The shift he saw in 2021 was in how software gets made. "Software development is actually a misnomer," he said. "We should really call it software assembly." Developers no longer write most of their code. They pull it from open-source repositories and, increasingly, generate it with AI models trained on code from across the internet. Some of it is excellent. Some of it is dangerous. Almost none of it comes from a source anyone has vetted. He compares it to a car. A manufacturer knows who supplied every part in a vehicle, where it came from, and how it was built. Software has no such discipline. A [Harvard Business School study](https://www.hbs.edu/ris/Publication%20Files/24-038%5F51f8444f-502c-4139-8bf2-56eb4b65c58a.pdf?ref=ceo.com) put the demand-side value of open-source software at $8.8 trillion, and some of the projects underneath that number have a single maintainer. "Sometimes that's literally one person working out of a garage somewhere," Badhwar said. Compromise that person's account, and the damage cascades through thousands of organizations at once. Attackers have figured this out. Why breach one company when you can poison a component used by all of them? The defensive math is worse than most boards understand. A typical large financial institution carries five to 10 million known security findings, Badhwar told the podcast. He estimates hundreds of billions of lines of code need to be rewritten, possibly trillions. No company employs enough developers for that. His answer is to turn AI on the problem it created: use it to isolate the small fraction of findings that pose real risk, then deploy agents that can read old codebases and automate the fixes. Endor Labs, which he founded with [Dimitri Stiliadis](https://www.endorlabs.com/learn/story-of-endorlabs?ref=ceo.com), raised a [$93 million Series B](https://www.endorlabs.com/learn/why-we-raised-a-93m-series-b-in-this-market?ref=ceo.com) led by DFJ Growth, bringing total funding to [$163 million](https://techcrunch.com/2025/04/23/endor-labs-which-builds-tools-to-scan-ai-generated-code-for-vulnerabilities-lands-93m/?ref=ceo.com). The company says ARR grew 30x over 18 months with 166% net revenue retention, and its platform now protects more than 5 million applications at customers including Rubrik and Mysten Labs. In February, it [acquired Autonomous Plane](https://www.prnewswire.com/news-releases/endor-labs-acquires-autonomous-plane-expanding-ai-native-application-security-with-full-stack-reachability-from-code-to-container-302684888.html?ref=ceo.com), a container security company founded by DockerSlim creator Kyle Quest. Some executives look at all this risk and conclude the answer is to slow down, keep everything on-premises, wait for the technology to settle. Badhwar watched companies try that with cloud computing. They lost. "Trying to slow down adoption isn't a viable strategy," he said. "The better approach is to embrace AI and build the right security controls around it." The same pragmatism shapes his view of the AI race with China, which he considers a misnomer. It is an infrastructure race. Compute requires data centers, data centers require power, and the capacity gets built wherever building happens fastest. He takes the objections to data centers seriously, from electricity demand to water consumption to strained grids, and expects the current phase of unlimited AI spending to end soon. One company he heard about burned through its annual AI budget in a few months. Optimization is coming. Running Endor Labs through all of this, Badhwar now spends roughly half his time on product, more than at any point in his career, because product-market fit shifts too fast to delegate. "One of the biggest mistakes organizations make is assuming that because something didn't work six months ago, it still won't work today," he said. "The willingness to unlearn and relearn has become a critical leadership skill." Every CEO.com interview ends with the same question: who gave you a chance? Badhwar named Vijay Jaju, a KPMG manager who put a 2006 computer science graduate on enterprise security projects at a time when almost nobody hired new graduates into the field. "That chance changed everything," he said. "We still stay in touch today." ```chapters 00:00:00 Introduction and Varun's Background 00:00:39 The Concept of Software Assembly 00:01:10 Security Challenges in Software Development 00:02:30 AI and Cybersecurity Vulnerabilities 00:03:08 Importance of Cybersecurity in the Age of AI 00:04:51 Understanding Software Supply Chain Security 00:08:36 Protecting Against Software Supply Chain Attacks 00:09:09 Open Source Security and OpenClaw 00:11:15 Embedding Security in Software Development 00:12:04 On-Prem vs. Cloud Security 00:13:49 The Scale of Code Rewriting Needed 00:17:34 Public Perception of AI and Data Centers 00:21:40 Token Costs and AI Utilization 00:22:50 The AI Race with China 00:25:15 A Day in the Life of a CEO 00:27:47 Core Principles of Leadership 00:30:19 The Importance of Giving and Taking Chances 00:31:43 Closing Remarks and Gratitude ``` ```vtt WEBVTT 00:00:00.240 --> 00:00:02.240 Varun,<00:00:00.880> thank<00:00:00.960> you<00:00:01.120> so<00:00:01.280> much<00:00:01.360> for<00:00:01.600> coming<00:00:01.760> on<00:00:01.920> the<00:00:02.000> show. 00:00:02.240 --> 00:00:04.880 You're<00:00:02.399> the<00:00:02.639> founder<00:00:02.879> and<00:00:03.120> CEO<00:00:03.520> of<00:00:03.760> Endor<00:00:04.319> Labs. 00:00:04.880 --> 00:00:08.400 Tell<00:00:04.960> us<00:00:05.120> what<00:00:05.279> Endor<00:00:05.680> Labs<00:00:06.000> is.<00:00:07.120> Yeah,<00:00:07.440> good<00:00:07.600> to<00:00:07.680> be<00:00:07.759> here,<00:00:08.000> Clint. 00:00:08.400 --> 00:00:12.320 So,<00:00:09.360> um,<00:00:10.160> my<00:00:10.400> quick<00:00:10.639> background<00:00:10.960> and<00:00:11.120> then<00:00:11.280> I'll<00:00:11.440> tell<00:00:11.519> you<00:00:11.679> about<00:00:11.840> the<00:00:12.000> company. 00:00:12.320 --> 00:00:19.359 I've<00:00:12.706> been<00:00:13.092> 16<00:00:13.478> years<00:00:13.864> of<00:00:14.249> building<00:00:14.635> cybersecurity<00:00:15.021> products,<00:00:15.407> uh,<00:00:15.793> at<00:00:16.179> the<00:00:16.565> intersection<00:00:16.951> of<00:00:17.336> mega<00:00:17.722> trends<00:00:18.108> and<00:00:18.494> technology<00:00:18.880> shifts. 00:00:19.359 --> 00:00:26.720 The<00:00:19.520> first<00:00:19.680> was<00:00:19.920> in<00:00:20.160> 2010.<00:00:21.279> Uh,<00:00:21.760> the<00:00:22.080> move<00:00:22.320> from<00:00:22.560> enterprise<00:00:23.039> applications<00:00:23.680> that<00:00:23.920> you<00:00:24.160> built<00:00:24.480> and<00:00:24.720> customized<00:00:25.199> to<00:00:25.519> SaaS. 00:00:26.720 --> 00:00:35.760 Uh,<00:00:27.053> 2015<00:00:27.387> was<00:00:27.720> the<00:00:28.053> move<00:00:28.386> from<00:00:28.720> data<00:00:29.053> center<00:00:29.386> to<00:00:29.720> cloud<00:00:30.053> platforms<00:00:30.386> like<00:00:30.720> AWS<00:00:31.053> and<00:00:31.386> I<00:00:31.719> built<00:00:32.053> the<00:00:32.386> cybersecurity<00:00:32.719> company<00:00:33.053> that<00:00:33.386> defined<00:00:33.719> that<00:00:34.052> category<00:00:34.386> called<00:00:34.719> Ardoq. 00:00:35.760 --> 00:00:44.000 And<00:00:35.920> in<00:00:36.160> 2021<00:00:36.960> I<00:00:37.120> started<00:00:37.360> Endor<00:00:37.840> Labs<00:00:38.160> because,<00:00:39.120> um,<00:00:39.280> my<00:00:39.600> thesis<00:00:40.000> was<00:00:40.800> software<00:00:41.280> development<00:00:42.239> is<00:00:42.559> actually<00:00:42.800> a<00:00:43.040> misnomer.<00:00:43.840> We 00:00:44.000 --> 00:00:52.079 should<00:00:44.079> really<00:00:44.239> call<00:00:44.480> it<00:00:44.640> software<00:00:44.960> assembly.<00:00:45.840> Most<00:00:46.079> people<00:00:46.320> don't<00:00:46.559> write<00:00:46.800> most<00:00:46.960> of<00:00:47.120> their<00:00:47.280> software<00:00:47.680> today;<00:00:48.079> they<00:00:48.640> assemble<00:00:49.120> components<00:00:49.680> from<00:00:50.000> open<00:00:50.320> source<00:00:51.520> and<00:00:51.760> certainly 00:00:52.079 --> 00:00:59.440 today<00:00:52.320> in<00:00:52.559> 2026<00:00:53.760> most<00:00:53.920> of<00:00:54.000> that<00:00:54.239> code<00:00:54.399> is<00:00:54.559> generated<00:00:54.879> by<00:00:55.039> AI<00:00:55.360> models<00:00:56.239> and<00:00:56.559> you're<00:00:56.800> just<00:00:56.960> kind<00:00:57.120> of<00:00:57.199> making<00:00:57.440> sure<00:00:57.520> that<00:00:57.760> it's<00:00:57.920> all<00:00:58.079> coming<00:00:58.320> together<00:00:58.640> as<00:00:58.879> an<00:00:59.039> application 00:00:59.440 --> 00:01:06.720 that<00:00:59.760> solves<00:01:00.000> a<00:01:00.239> business<00:01:00.399> need<00:01:00.640> for<00:01:00.879> you.<00:01:01.520> And<00:01:01.680> so<00:01:01.840> our<00:01:02.000> core<00:01:02.239> thesis<00:01:02.559> starting<00:01:02.800> Endor<00:01:03.199> Labs<00:01:03.520> was<00:01:04.080> you're<00:01:04.320> going<00:01:04.400> to<00:01:04.559> have<00:01:04.960> more<00:01:05.280> people<00:01:05.600> producing<00:01:06.000> more<00:01:06.320> software 00:01:06.720 --> 00:01:10.960 than<00:01:06.960> ever<00:01:07.520> and<00:01:07.840> most<00:01:08.080> of<00:01:08.240> it<00:01:08.799> emerges<00:01:09.280> from<00:01:09.520> untrusted<00:01:10.320> sources. 00:01:10.960 --> 00:01:18.400 It<00:01:11.200> comes<00:01:11.439> from<00:01:12.080> Stack<00:01:12.400> Overflow<00:01:13.040> and<00:01:13.360> github.com,<00:01:14.479> which<00:01:15.040> who<00:01:15.360> knows<00:01:15.520> who<00:01:15.680> built<00:01:15.920> the<00:01:16.080> software,<00:01:16.799> and<00:01:16.960> it<00:01:17.200> just<00:01:17.360> kind<00:01:17.439> of<00:01:17.520> exists<00:01:17.920> there. 00:01:18.400 --> 00:01:26.560 And<00:01:18.560> now,<00:01:18.799> lo<00:01:18.960> and<00:01:19.040> behold,<00:01:19.520> the<00:01:19.920> models<00:01:20.320> and<00:01:20.560> LLMs<00:01:21.840> are,<00:01:22.880> uh,<00:01:23.119> training<00:01:23.600> how<00:01:23.840> to<00:01:23.920> write<00:01:24.240> software<00:01:24.640> based<00:01:24.960> on<00:01:25.200> this<00:01:25.840> software<00:01:26.159> that<00:01:26.400> has 00:01:26.560 --> 00:01:32.720 existed<00:01:26.880> on<00:01:27.040> the<00:01:27.200> internet<00:01:27.520> that<00:01:28.080> a<00:01:28.240> lot<00:01:28.320> of<00:01:28.479> it<00:01:28.640> is<00:01:28.799> good,<00:01:29.439> but<00:01:29.759> a<00:01:30.000> lot<00:01:30.080> of<00:01:30.159> it<00:01:30.320> has<00:01:30.479> a<00:01:30.640> lot<00:01:30.720> of<00:01:30.880> security<00:01:31.280> problems. 00:01:32.720 --> 00:01:41.600 And<00:01:33.040> the<00:01:33.280> thesis<00:01:33.680> was<00:01:34.479> that<00:01:34.720> there<00:01:34.880> is<00:01:35.119> always<00:01:35.520> this<00:01:36.000> contention<00:01:36.799> and<00:01:37.119> trade-off<00:01:37.759> between<00:01:38.640> can<00:01:38.880> I<00:01:39.040> ship<00:01:39.439> fast<00:01:39.920> as<00:01:40.159> a<00:01:40.400> business<00:01:41.200> and 00:01:41.600 --> 00:01:49.200 be<00:01:41.759> competitive<00:01:42.320> and<00:01:42.640> have<00:01:42.880> software<00:01:43.360> that<00:01:43.600> is<00:01:43.759> coming<00:01:43.920> out<00:01:44.159> and<00:01:44.400> features<00:01:44.720> that<00:01:44.880> are<00:01:45.040> coming<00:01:45.200> out<00:01:45.439> continuously,<00:01:46.880> or<00:01:47.759> can<00:01:48.000> I<00:01:48.159> ship<00:01:48.320> it<00:01:48.560> securely? 00:01:49.200 --> 00:01:57.119 It<00:01:49.360> was<00:01:49.439> an<00:01:49.600> or<00:01:50.079> because<00:01:50.640> the<00:01:50.960> cost<00:01:51.200> of<00:01:51.360> doing<00:01:51.600> security<00:01:52.079> well<00:01:52.320> as<00:01:52.479> you<00:01:52.640> develop<00:01:52.960> software<00:01:53.360> was<00:01:53.520> really<00:01:53.759> high<00:01:54.640> and<00:01:54.880> so<00:01:55.200> people<00:01:55.439> missed<00:01:55.759> those<00:01:56.000> steps<00:01:56.320> and<00:01:56.479> and<00:01:56.720> took<00:01:56.960> some 00:01:57.119 --> 00:02:03.360 shortcuts<00:01:58.159> and<00:01:58.479> we<00:01:58.640> wanted<00:01:58.880> to<00:01:59.040> make<00:01:59.200> it<00:01:59.360> an<00:01:59.520> and<00:02:00.000> that<00:02:00.240> you<00:02:00.560> can<00:02:00.799> ship<00:02:01.200> fast<00:02:01.920> and<00:02:02.320> you<00:02:02.479> can<00:02:02.640> ship<00:02:02.960> securely. 00:02:03.360 --> 00:02:11.200 That<00:02:03.692> was<00:02:04.024> a<00:02:04.356> North<00:02:04.688> Star:<00:02:05.020> is<00:02:05.352> how<00:02:05.684> do<00:02:06.016> we<00:02:06.348> make<00:02:06.680> development<00:02:07.012> the<00:02:07.344> focus<00:02:07.676> while<00:02:08.008> it's<00:02:08.340> all<00:02:08.672> just<00:02:09.004> secure<00:02:09.336> as<00:02:09.668> a<00:02:10.000> byproduct. 00:02:11.200 --> 00:02:19.440 Um,<00:02:11.920> and<00:02:12.160> you<00:02:12.319> know,<00:02:12.400> we<00:02:12.560> were<00:02:12.720> kind<00:02:12.879> of<00:02:12.959> talking<00:02:13.120> a<00:02:13.360> lot<00:02:13.440> about<00:02:13.680> this,<00:02:14.640> uh,<00:02:14.720> and<00:02:14.959> obviously<00:02:15.360> today<00:02:15.599> we<00:02:15.840> sit<00:02:16.000> in<00:02:16.160> a<00:02:16.400> world<00:02:16.560> where,<00:02:17.520> uh,<00:02:17.680> everybody's<00:02:18.239> conversation<00:02:18.720> on<00:02:18.959> AI<00:02:19.280> and 00:02:19.440 --> 00:02:27.120 LLM<00:02:19.823> writing<00:02:20.206> code<00:02:20.589> is<00:02:20.973> cybersecurity<00:02:21.356> and<00:02:21.739> the<00:02:22.122> challenges<00:02:22.505> of<00:02:22.888> every<00:02:23.272> security<00:02:23.655> vulnerability<00:02:24.038> is<00:02:24.421> going<00:02:24.804> to<00:02:25.187> be<00:02:25.571> easily<00:02:25.954> found<00:02:26.337> by<00:02:26.720> AI. 00:02:27.120 --> 00:02:30.720 So<00:02:27.280> what<00:02:27.360> are<00:02:27.520> we<00:02:27.599> going<00:02:27.680> to<00:02:27.760> do?<00:02:27.920> And<00:02:28.080> so<00:02:28.239> here<00:02:28.400> we<00:02:28.560> are.<00:02:28.720> I'm<00:02:28.959> looking<00:02:29.040> forward<00:02:29.280> to<00:02:29.440> this<00:02:29.599> conversation. 00:02:30.720 --> 00:02:33.360 Well,<00:02:30.959> yeah,<00:02:31.200> we're<00:02:31.440> talking<00:02:31.680> at<00:02:31.840> a<00:02:32.080> really<00:02:32.319> interesting<00:02:32.720> time,<00:02:33.040> right? 00:02:33.360 --> 00:02:41.040 Like,<00:02:33.760> uh,<00:02:33.920> Anthropic<00:02:34.640> says<00:02:34.800> they<00:02:35.040> have<00:02:35.120> this<00:02:35.440> model<00:02:35.760> called<00:02:36.080> Mythos,<00:02:37.120> um,<00:02:37.920> that<00:02:38.319> they<00:02:38.560> don't<00:02:38.720> even<00:02:38.879> dare<00:02:39.200> release<00:02:39.599> publicly 00:02:41.040 --> 00:02:44.400 because<00:02:41.280> of<00:02:41.440> the<00:02:41.680> vulnerabilities<00:02:42.400> it's<00:02:42.640> finding<00:02:43.200> in<00:02:43.519> software. 00:02:44.400 --> 00:02:50.640 I<00:02:44.720> assume,<00:02:45.599> um,<00:02:46.319> one<00:02:46.480> or<00:02:46.640> two<00:02:46.800> of<00:02:46.879> the<00:02:46.959> other<00:02:47.120> frontier<00:02:47.599> labs<00:02:47.920> might<00:02:48.160> have<00:02:48.400> something<00:02:48.720> similar<00:02:49.360> already. 00:02:50.640 --> 00:02:53.280 Um,<00:02:52.400> let's<00:02:52.720> start<00:02:52.879> there. 00:02:53.280 --> 00:02:55.599 I<00:02:53.584> mean,<00:02:53.888> how<00:02:54.192> important<00:02:54.496> is<00:02:54.800> cybersecurity? 00:02:55.599 --> 00:03:02.480 I<00:02:55.680> mean,<00:02:55.840> it's<00:02:56.000> always<00:02:56.239> been<00:02:56.400> super<00:02:56.720> important,<00:02:57.440> but<00:02:57.680> man,<00:02:58.080> I<00:02:58.239> can't<00:02:58.480> even<00:02:58.640> think<00:02:58.800> of<00:02:58.959> like<00:02:59.040> a<00:02:59.280> more<00:02:59.440> important<00:02:59.840> thing<00:03:00.080> given<00:03:00.400> the<00:03:00.560> age<00:03:00.800> of<00:03:00.879> AI<00:03:01.200> and<00:03:01.440> given<00:03:01.760> things<00:03:01.920> like<00:03:02.159> cloud 00:03:02.480 --> 00:03:08.400 Mythos<00:03:02.879> and<00:03:03.120> things<00:03:03.200> that<00:03:03.360> we're<00:03:03.599> hearing<00:03:04.319> and<00:03:04.640> the<00:03:04.800> vulnerabilities<00:03:05.680> that<00:03:05.840> AI<00:03:06.239> can<00:03:06.400> exploit. 00:03:08.400 --> 00:03:11.120 Yeah.<00:03:08.720> Great.<00:03:08.879> And<00:03:09.120> look,<00:03:09.440> this<00:03:09.599> is<00:03:09.840> kind<00:03:10.000> of<00:03:10.080> going<00:03:10.319> back<00:03:10.480> to<00:03:10.640> what<00:03:10.800> I<00:03:10.959> said. 00:03:11.120 --> 00:03:13.760 It<00:03:11.280> was,<00:03:11.599> you<00:03:11.760> know,<00:03:11.920> ship<00:03:12.319> fast<00:03:12.720> or<00:03:13.040> ship<00:03:13.360> secure. 00:03:13.760 --> 00:03:21.120 And<00:03:13.920> so<00:03:14.080> a<00:03:14.239> lot<00:03:14.319> of<00:03:14.400> times<00:03:14.720> people<00:03:15.200> chose<00:03:15.599> ship<00:03:16.000> fast<00:03:16.480> because<00:03:16.720> you<00:03:16.879> kind<00:03:16.959> of<00:03:17.040> had<00:03:17.200> a<00:03:17.360> hope<00:03:17.599> and<00:03:17.760> pray<00:03:18.080> strategy<00:03:19.200> that<00:03:19.519> look,<00:03:19.840> yeah,<00:03:20.159> there<00:03:20.319> might<00:03:20.560> be<00:03:20.640> some<00:03:20.800> problems, 00:03:21.120 --> 00:03:23.599 but<00:03:21.360> what<00:03:21.519> are<00:03:21.680> the<00:03:21.760> chances<00:03:22.159> somebody<00:03:22.480> finds<00:03:22.879> them? 00:03:23.599 --> 00:03:30.959 And<00:03:23.840> today<00:03:24.159> with<00:03:24.319> how<00:03:24.640> good<00:03:24.800> the<00:03:25.440> LLMs<00:03:25.920> have<00:03:26.080> gotten,<00:03:26.720> they<00:03:27.040> will<00:03:27.360> find<00:03:27.680> every<00:03:28.080> security<00:03:28.480> issue<00:03:28.800> that<00:03:29.040> is<00:03:29.280> hidden<00:03:29.519> in<00:03:29.760> your<00:03:30.000> software. 00:03:30.959 --> 00:03:36.080 Um,<00:03:31.680> kudos<00:03:32.159> to<00:03:32.239> Anthropic,<00:03:32.879> like<00:03:33.040> they're<00:03:33.280> a<00:03:33.440> tremendous<00:03:34.000> innovation<00:03:34.560> and<00:03:34.799> marketing<00:03:35.280> machine. 00:03:36.080 --> 00:03:44.319 Uh,<00:03:36.239> Mythos<00:03:36.720> is<00:03:36.959> probably<00:03:37.280> fantastic<00:03:37.840> and<00:03:38.080> people<00:03:38.400> I<00:03:38.640> talked<00:03:38.799> to,<00:03:38.959> a<00:03:39.120> lot<00:03:39.200> of<00:03:39.360> organizations<00:03:39.920> that<00:03:40.080> have<00:03:40.239> early<00:03:40.560> access<00:03:40.799> to<00:03:40.879> it<00:03:41.200> or<00:03:41.360> using<00:03:41.599> it,<00:03:42.319> no<00:03:42.560> doubt<00:03:42.879> is<00:03:43.120> great,<00:03:44.080> but 00:03:44.319 --> 00:03:52.640 to<00:03:44.560> your<00:03:44.879> point,<00:03:46.080> OpenAI's<00:03:46.720> Daybreak<00:03:47.200> is<00:03:47.440> not<00:03:47.599> that<00:03:47.840> far<00:03:48.319> behind<00:03:48.959> and<00:03:49.360> that's<00:03:49.840> mostly<00:03:50.319> available<00:03:51.280> where<00:03:51.519> it's<00:03:51.840> not<00:03:52.000> that<00:03:52.239> far<00:03:52.400> from 00:03:52.640 --> 00:03:59.760 GPT-5<00:03:53.680> that's<00:03:54.000> available.<00:03:54.959> Heck,<00:03:55.200> I<00:03:55.360> would<00:03:55.519> go<00:03:55.599> as<00:03:55.840> far<00:03:55.920> as<00:03:56.080> saying<00:03:56.239> like<00:03:56.480> a<00:03:56.640> lot<00:03:56.720> of<00:03:56.799> open<00:03:57.040> source<00:03:57.280> models<00:03:57.599> emerging<00:03:57.920> out<00:03:58.080> of<00:03:58.239> China<00:03:59.040> are<00:03:59.280> pretty<00:03:59.519> good. 00:03:59.760 --> 00:04:08.400 The<00:03:59.840> the<00:03:59.920> the<00:04:00.400> funny<00:04:00.640> thing<00:04:00.799> is<00:04:01.760> as<00:04:02.080> an<00:04:02.959> organization<00:04:03.760> that<00:04:04.080> is<00:04:04.239> trying<00:04:04.400> to<00:04:04.560> play<00:04:04.799> defense<00:04:05.280> against<00:04:05.680> all<00:04:05.840> of<00:04:06.000> these<00:04:06.319> things,<00:04:07.120> you<00:04:07.280> got<00:04:07.360> to<00:04:07.519> be<00:04:07.599> right<00:04:07.920> 100% 00:04:08.400 --> 00:04:11.599 of<00:04:08.560> the<00:04:08.720> time,<00:04:08.959> meaning<00:04:09.200> you<00:04:09.360> have<00:04:09.439> to<00:04:09.599> defend<00:04:10.319> 100%<00:04:10.799> of<00:04:10.879> the<00:04:11.120> time. 00:04:11.599 --> 00:04:14.720 The<00:04:11.840> attackers<00:04:12.400> don't<00:04:12.560> need<00:04:12.720> to<00:04:12.879> find<00:04:13.200> every<00:04:13.519> last<00:04:13.840> bug<00:04:14.000> in<00:04:14.239> your<00:04:14.319> software. 00:04:14.720 --> 00:04:18.639 They<00:04:14.879> just<00:04:15.040> need<00:04:15.120> to<00:04:15.280> find<00:04:15.519> one<00:04:16.560> that<00:04:16.799> you<00:04:17.040> haven't<00:04:17.280> fixed<00:04:17.600> and<00:04:17.840> find<00:04:18.079> your<00:04:18.239> way<00:04:18.320> in. 00:04:18.639 --> 00:04:21.040 And<00:04:18.799> there's<00:04:19.040> plenty<00:04:19.280> of<00:04:19.440> low-hanging<00:04:20.000> fruit,<00:04:20.320> candidly. 00:04:21.040 --> 00:04:29.360 So<00:04:21.199> I<00:04:21.359> think<00:04:21.440> Mythos<00:04:22.000> really<00:04:22.160> sets<00:04:22.479> the<00:04:22.639> tone<00:04:22.960> for,<00:04:24.160> hey,<00:04:24.479> like<00:04:24.720> we<00:04:24.960> can<00:04:25.199> no<00:04:25.440> longer<00:04:26.400> ship<00:04:26.880> software<00:04:27.680> with<00:04:27.919> security<00:04:28.320> vulnerabilities<00:04:29.120> because 00:04:29.360 --> 00:04:32.720 it<00:04:29.680> will<00:04:29.919> get<00:04:30.080> found.<00:04:30.639> I<00:04:30.800> think<00:04:30.880> it's<00:04:31.120> a<00:04:31.360> complete<00:04:31.840> mindset<00:04:32.320> shift. 00:04:32.720 --> 00:04:36.960 It's<00:04:32.960> a<00:04:33.120> complete<00:04:33.440> rewrite<00:04:34.240> of<00:04:34.479> how<00:04:34.720> we've<00:04:34.880> thought<00:04:35.120> about<00:04:35.280> software<00:04:35.840> security. 00:04:36.960 --> 00:04:42.479 It<00:04:37.199> is<00:04:37.360> dangerous,<00:04:37.759> but<00:04:37.919> just<00:04:38.160> because<00:04:38.560> Anthropic<00:04:39.199> hasn't<00:04:39.440> released<00:04:39.759> it<00:04:39.919> doesn't<00:04:40.160> make<00:04:40.240> us<00:04:40.400> any<00:04:40.639> safer<00:04:41.040> today. 00:04:42.479 --> 00:04:45.199 Could<00:04:42.639> you<00:04:42.800> talk<00:04:42.960> about<00:04:43.120> like<00:04:43.280> the<00:04:43.600> software<00:04:44.080> supply<00:04:44.560> chain? 00:04:45.199 --> 00:04:48.800 I've<00:04:45.360> heard<00:04:45.520> you<00:04:45.759> talk<00:04:45.840> about<00:04:46.000> this<00:04:46.240> before<00:04:46.479> and<00:04:46.800> like<00:04:47.040> securing<00:04:47.520> the<00:04:47.840> software<00:04:48.240> supply<00:04:48.560> chain. 00:04:48.800 --> 00:04:55.919 Give<00:04:48.880> us<00:04:49.040> a<00:04:49.120> sense<00:04:49.280> by<00:04:49.600> what<00:04:49.759> you<00:04:49.919> mean<00:04:50.080> by<00:04:50.240> that.<00:04:51.120> Yeah,<00:04:51.360> Clint,<00:04:51.600> so<00:04:51.759> going<00:04:52.000> back<00:04:52.080> to<00:04:52.240> what<00:04:52.400> I<00:04:52.479> said,<00:04:52.720> right,<00:04:52.960> which<00:04:53.120> is<00:04:53.360> developers<00:04:54.320> assemble<00:04:54.880> software,<00:04:55.440> not<00:04:55.600> write 00:04:55.919 --> 00:04:58.560 software.<00:04:56.240> It's<00:04:56.400> a<00:04:56.560> bit<00:04:56.639> like<00:04:56.800> a<00:04:57.120> car<00:04:57.440> manufacturing<00:04:58.160> line. 00:04:58.560 --> 00:05:00.960 All<00:04:58.720> the<00:04:58.960> parts<00:04:59.199> are<00:04:59.360> built<00:04:59.600> by<00:04:59.759> lots<00:05:00.080> of<00:05:00.160> other<00:05:00.400> factories. 00:05:00.960 --> 00:05:08.800 Your<00:05:01.120> job<00:05:01.280> is<00:05:01.440> to<00:05:01.600> put<00:05:01.759> it<00:05:01.919> together.<00:05:03.040> Except<00:05:03.360> when<00:05:03.520> it's<00:05:03.680> a<00:05:03.919> car<00:05:04.080> assembly<00:05:04.560> line,<00:05:05.199> you<00:05:05.360> know,<00:05:05.600> every<00:05:05.919> manufacturer,<00:05:06.960> you<00:05:07.120> have<00:05:07.199> a<00:05:07.360> bill<00:05:07.520> of<00:05:07.759> material<00:05:08.000> of<00:05:08.240> when<00:05:08.479> it<00:05:08.639> was 00:05:08.800 --> 00:05:13.199 built,<00:05:09.120> how<00:05:09.360> it<00:05:09.520> was<00:05:09.680> built,<00:05:10.400> what<00:05:10.560> are<00:05:10.639> the<00:05:10.800> ingredients<00:05:11.199> in<00:05:11.440> that<00:05:11.759> component. 00:05:13.199 --> 00:05:19.120 Um,<00:05:13.919> software<00:05:14.400> supply<00:05:14.720> chain<00:05:15.120> security<00:05:15.520> is<00:05:15.759> the<00:05:16.000> idea<00:05:16.960> that<00:05:17.680> the<00:05:17.919> same<00:05:18.080> thing<00:05:18.160> is<00:05:18.320> happening<00:05:18.560> with<00:05:18.720> software. 00:05:19.120 --> 00:05:22.639 When<00:05:19.280> I<00:05:19.440> say<00:05:19.520> I'm<00:05:19.680> going<00:05:19.759> to<00:05:19.840> build<00:05:20.000> an<00:05:20.240> application<00:05:20.639> that<00:05:20.880> does<00:05:21.039> A,<00:05:21.280> B,<00:05:21.440> or<00:05:21.680> C, 00:05:22.639 --> 00:05:30.080 I'm<00:05:22.960> bringing<00:05:23.280> components<00:05:23.759> of<00:05:23.919> a<00:05:24.080> database<00:05:24.479> that<00:05:24.720> is<00:05:24.880> open<00:05:25.120> source,<00:05:25.600> a<00:05:25.919> logging<00:05:26.400> service<00:05:26.720> that<00:05:26.960> is<00:05:27.039> open<00:05:27.360> source,<00:05:28.080> another<00:05:28.479> component,<00:05:28.880> and<00:05:29.039> then<00:05:29.199> AI<00:05:29.520> is<00:05:29.680> putting<00:05:29.919> it 00:05:30.080 --> 00:05:34.080 all<00:05:30.320> together<00:05:31.039> by<00:05:31.360> coding<00:05:31.680> the<00:05:32.000> layer<00:05:32.160> that<00:05:32.320> brings<00:05:32.560> all<00:05:32.720> of<00:05:32.880> this<00:05:33.039> together. 00:05:34.080 --> 00:05:39.280 But<00:05:34.479> 90%<00:05:35.199> of<00:05:35.360> the<00:05:35.680> things<00:05:35.919> in<00:05:36.160> this<00:05:36.479> application<00:05:37.520> are<00:05:37.759> not<00:05:38.080> created<00:05:38.479> by<00:05:38.720> you. 00:05:39.280 --> 00:05:46.240 They're<00:05:39.600> created<00:05:39.919> by<00:05:40.160> somebody<00:05:40.479> else.<00:05:41.919> Except<00:05:42.400> unlike<00:05:42.880> car<00:05:43.199> manufacturing,<00:05:44.240> we<00:05:44.479> have<00:05:44.639> no<00:05:44.880> idea<00:05:45.120> who<00:05:45.360> built<00:05:45.600> this. 00:05:46.240 --> 00:05:48.479 We<00:05:46.479> don't<00:05:46.560> know<00:05:46.639> what<00:05:46.880> their<00:05:47.120> motivations<00:05:47.759> are. 00:05:48.479 --> 00:05:53.039 Can<00:05:48.720> we<00:05:48.960> trust<00:05:49.280> them?<00:05:50.160> Are<00:05:50.400> they<00:05:50.560> going<00:05:50.639> to<00:05:50.800> stand<00:05:51.039> behind<00:05:51.280> the<00:05:51.520> quality<00:05:51.840> and<00:05:52.080> security<00:05:52.320> of<00:05:52.479> their<00:05:52.639> product? 00:05:53.039 --> 00:06:02.080 We<00:05:53.199> just<00:05:53.360> don't<00:05:53.600> know.<00:05:54.320> And<00:05:54.560> so<00:05:54.880> there<00:05:55.120> has<00:05:55.280> been<00:05:55.440> this<00:05:55.759> implicit<00:05:56.400> trust<00:05:57.039> for<00:05:57.280> the<00:05:57.440> last<00:05:57.600> 20,<00:05:57.919> 30<00:05:58.240> years<00:05:58.479> around<00:05:58.800> a<00:05:59.120> very<00:05:59.520> powerful<00:05:59.919> ecosystem<00:06:00.479> of<00:06:00.639> open<00:06:00.960> source<00:06:01.280> software. 00:06:02.080 --> 00:06:09.680 In<00:06:02.240> fact,<00:06:02.400> there's<00:06:02.560> a<00:06:02.720> Harvard<00:06:03.120> study<00:06:03.759> that<00:06:04.080> calls<00:06:04.319> it<00:06:04.560> a<00:06:04.880> trillion<00:06:05.520> dollar<00:06:05.840> economic<00:06:06.479> impact<00:06:07.440> of,<00:06:08.160> um,<00:06:08.880> of<00:06:09.120> software. 00:06:09.680 --> 00:06:18.400 Phenomenal,<00:06:10.479> right?<00:06:11.440> But<00:06:12.000> at<00:06:12.160> the<00:06:12.319> same<00:06:12.560> time,<00:06:13.520> we<00:06:13.759> trust<00:06:14.080> complete<00:06:14.479> strangers<00:06:14.960> on<00:06:15.120> the<00:06:15.280> internet<00:06:15.680> to<00:06:15.919> provide<00:06:16.319> components<00:06:17.199> of<00:06:17.600> software<00:06:18.160> that 00:06:18.400 --> 00:06:21.039 are<00:06:18.639> mission-critical<00:06:19.280> to<00:06:19.440> us<00:06:19.680> and<00:06:19.840> our<00:06:20.000> organization. 00:06:21.039 --> 00:06:28.800 And<00:06:21.280> there<00:06:21.440> are<00:06:21.520> tremendous<00:06:22.080> cases<00:06:22.479> out<00:06:22.639> there<00:06:22.880> that<00:06:23.120> you<00:06:23.280> can<00:06:23.440> read<00:06:23.600> about<00:06:24.479> where<00:06:24.720> it<00:06:25.039> turns<00:06:25.199> out<00:06:25.360> that<00:06:25.600> this<00:06:25.840> component<00:06:26.240> was<00:06:26.400> built<00:06:26.639> by<00:06:26.800> a<00:06:26.960> guy<00:06:27.120> in<00:06:27.280> a<00:06:27.440> garage<00:06:27.680> in<00:06:27.919> Russia 00:06:28.800 --> 00:06:36.800 that<00:06:29.120> we<00:06:29.280> don't<00:06:29.440> even<00:06:29.600> know.<00:06:29.759> And<00:06:29.919> he's<00:06:30.160> pleading<00:06:30.560> on<00:06:30.720> the<00:06:30.880> internet<00:06:31.440> that<00:06:31.680> he<00:06:31.919> has<00:06:32.160> no<00:06:32.319> money,<00:06:32.639> he's<00:06:32.800> been<00:06:32.960> to<00:06:33.120> jail,<00:06:33.680> he's<00:06:33.919> in<00:06:34.160> debt,<00:06:34.880> and<00:06:35.120> now<00:06:35.360> his<00:06:35.680> software<00:06:36.080> is<00:06:36.319> a<00:06:36.479> component 00:06:36.800 --> 00:06:44.960 of<00:06:36.960> the<00:06:37.199> internet.<00:06:38.319> And<00:06:38.800> if<00:06:39.199> that<00:06:39.600> person<00:06:40.400> or<00:06:40.720> that<00:06:41.199> piece<00:06:41.520> of<00:06:41.759> software<00:06:42.160> that<00:06:42.479> they<00:06:42.639> had<00:06:43.199> as<00:06:43.520> just<00:06:43.680> a<00:06:43.840> good<00:06:44.080> Samaritan<00:06:44.639> put<00:06:44.800> out 00:06:44.960 --> 00:06:53.680 on<00:06:45.120> the<00:06:45.280> internet,<00:06:45.919> somebody<00:06:46.240> chooses<00:06:46.639> to<00:06:46.880> take<00:06:47.039> over<00:06:47.520> or<00:06:47.840> pay<00:06:48.080> them<00:06:48.240> off<00:06:48.400> and<00:06:48.639> take<00:06:48.880> the<00:06:49.120> the<00:06:49.440> account<00:06:49.840> or<00:06:50.639> kind<00:06:50.880> of<00:06:50.960> breach<00:06:51.360> their<00:06:51.680> access<00:06:52.160> and<00:06:52.479> take<00:06:52.720> over. 00:06:53.680 --> 00:06:59.680 The<00:06:53.919> ripple<00:06:54.400> down<00:06:54.560> effect<00:06:55.039> is<00:06:55.440> like<00:06:56.080> hundreds<00:06:56.479> of<00:06:56.639> thousands<00:06:57.039> of<00:06:57.440> organizations<00:06:58.160> impacted<00:06:58.800> in<00:06:58.960> a<00:06:59.199> day. 00:06:59.680 --> 00:07:02.960 That<00:06:59.919> is<00:07:00.160> essentially<00:07:00.560> software<00:07:00.960> supply<00:07:01.199> chain<00:07:01.599> security,<00:07:02.400> right? 00:07:02.960 --> 00:07:11.599 It<00:07:03.311> turtles<00:07:03.663> all<00:07:04.014> the<00:07:04.365> way<00:07:04.716> down,<00:07:05.068> like<00:07:05.419> the<00:07:05.770> most<00:07:06.121> powerful<00:07:06.473> critical,<00:07:06.824> mission-critical<00:07:07.175> systems<00:07:07.526> rely<00:07:07.878> on<00:07:08.229> some<00:07:08.580> software<00:07:08.931> built<00:07:09.283> by<00:07:09.634> some<00:07:09.985> person<00:07:10.336> in<00:07:10.688> their<00:07:11.039> garage 00:07:11.599 --> 00:07:14.960 as<00:07:11.840> a<00:07:12.000> best<00:07:12.240> effort<00:07:12.479> that<00:07:12.720> we<00:07:12.880> have<00:07:13.039> no<00:07:13.199> contractual<00:07:13.759> obligations<00:07:14.400> with. 00:07:14.960 --> 00:07:23.120 And<00:07:15.324> now<00:07:15.687> what's<00:07:16.051> happening,<00:07:16.415> the<00:07:16.778> way<00:07:17.142> attackers<00:07:17.505> work<00:07:17.869> and<00:07:18.233> cybersecurity<00:07:18.596> works<00:07:18.960> is<00:07:19.324> the<00:07:19.687> attackers<00:07:20.051> are<00:07:20.415> looking<00:07:20.778> for<00:07:21.142> the<00:07:21.505> easiest<00:07:21.869> way<00:07:22.233> in<00:07:22.596> that<00:07:22.960> can 00:07:23.120 --> 00:07:28.080 have<00:07:23.280> a<00:07:23.440> high<00:07:23.680> blast<00:07:24.080> radius<00:07:24.479> across<00:07:24.800> the<00:07:24.960> industry<00:07:25.759> and<00:07:26.000> today<00:07:26.639> that<00:07:26.880> is<00:07:27.039> open<00:07:27.360> source. 00:07:28.080 --> 00:07:36.800 They<00:07:28.400> are<00:07:28.560> finding<00:07:28.960> every<00:07:29.360> single<00:07:29.680> day<00:07:30.000> ways<00:07:30.720> to<00:07:30.960> breach<00:07:31.759> one<00:07:32.080> account<00:07:32.720> or<00:07:33.039> one<00:07:33.280> piece<00:07:33.520> of<00:07:33.680> software<00:07:34.400> that<00:07:34.639> is<00:07:34.800> used<00:07:35.039> by<00:07:35.280> 32<00:07:35.759> million<00:07:36.080> other 00:07:36.800 --> 00:07:39.840 companies<00:07:37.199> and<00:07:37.440> and<00:07:37.520> and<00:07:38.319> applications. 00:07:39.840 --> 00:07:44.000 Compromise<00:07:40.319> that<00:07:40.560> one<00:07:40.800> little<00:07:41.039> piece<00:07:41.199> of<00:07:41.360> software<00:07:41.680> that<00:07:41.919> might<00:07:42.000> be<00:07:42.160> 10<00:07:42.319> lines<00:07:42.560> of<00:07:42.720> code. 00:07:44.000 --> 00:07:50.720 And<00:07:44.160> there<00:07:44.400> you<00:07:44.560> go.<00:07:44.800> Now<00:07:45.039> you've<00:07:45.280> kind<00:07:45.440> of<00:07:46.160> like,<00:07:46.400> instead<00:07:46.639> of<00:07:46.800> going<00:07:46.960> after<00:07:47.199> one<00:07:47.440> organization<00:07:48.000> at<00:07:48.160> a<00:07:48.319> time,<00:07:48.800> I've<00:07:49.120> gone<00:07:49.280> after<00:07:49.599> thousands. 00:07:50.720 --> 00:07:54.479 And<00:07:50.880> that's<00:07:51.039> a<00:07:51.280> scary<00:07:51.599> fact.<00:07:52.319> And<00:07:52.560> we've<00:07:52.880> never<00:07:53.199> had<00:07:53.440> defenses<00:07:53.919> against<00:07:54.160> it. 00:07:54.479 --> 00:07:58.639 Believe<00:07:54.720> it<00:07:54.800> or<00:07:54.960> not,<00:07:55.360> it's<00:07:55.520> been<00:07:55.840> 100%<00:07:56.479> trust<00:07:57.120> on<00:07:57.360> this<00:07:57.599> ecosystem. 00:07:58.639 --> 00:08:05.039 And<00:07:58.879> with<00:07:59.039> software<00:07:59.440> supply<00:07:59.680> chain<00:07:59.919> security,<00:08:00.240> the<00:08:00.479> whole<00:08:00.639> idea<00:08:00.960> coming<00:08:01.199> to<00:08:01.360> the<00:08:01.440> forefront<00:08:01.919> today<00:08:02.720> because<00:08:02.960> of<00:08:03.120> breaches<00:08:03.599> happening<00:08:03.840> every<00:08:04.160> single<00:08:04.400> day<00:08:04.639> resulting 00:08:05.039 --> 00:08:09.680 from<00:08:05.199> this.<00:08:05.440> In<00:08:05.599> fact,<00:08:06.400> uh,<00:08:06.560> last<00:08:06.879> week,<00:08:07.199> GitHub<00:08:07.759> disclosed<00:08:08.560> that<00:08:08.879> they<00:08:09.120> got<00:08:09.280> breached. 00:08:09.680 --> 00:08:19.120 The<00:08:09.759> entirety<00:08:10.080> of<00:08:10.240> github.com,<00:08:10.960> a<00:08:11.199> Microsoft<00:08:11.759> company,<00:08:12.800> got<00:08:13.039> breached<00:08:13.360> because<00:08:13.599> one<00:08:13.840> developer<00:08:15.039> downloaded<00:08:15.680> one<00:08:16.080> piece<00:08:16.479> of<00:08:16.639> an<00:08:16.879> extension<00:08:17.360> to<00:08:17.599> their<00:08:17.840> IDE 00:08:19.120 --> 00:08:24.800 that<00:08:19.440> was<00:08:19.599> malicious.<00:08:20.800> So,<00:08:21.360> you<00:08:21.520> know,<00:08:21.680> these<00:08:21.919> developer<00:08:22.319> machines<00:08:22.639> are<00:08:23.120> incredibly<00:08:23.759> powerful. 00:08:24.800 --> 00:08:33.120 Our<00:08:25.305> reliance<00:08:25.810> on<00:08:26.315> open-source<00:08:26.820> software<00:08:27.325> is<00:08:27.830> extremely<00:08:28.335> high<00:08:28.840> and<00:08:29.345> our<00:08:29.850> guardrails<00:08:30.355> to<00:08:30.860> defend<00:08:31.365> against<00:08:31.870> attack<00:08:32.375> in<00:08:32.880> this 00:08:33.120 --> 00:08:40.320 ecosystem<00:08:33.760> are<00:08:34.080> near<00:08:34.320> zero.<00:08:36.000> So,<00:08:36.240> it<00:08:36.479> sounds<00:08:36.640> like<00:08:36.800> advantage<00:08:37.279> hacker,<00:08:38.800> huh? 00:08:40.320 --> 00:08:43.279 So,<00:08:40.959> so<00:08:41.279> how<00:08:41.440> do<00:08:41.599> we<00:08:41.839> how<00:08:42.080> do<00:08:42.159> you<00:08:42.399> protect<00:08:42.640> against<00:08:42.959> this? 00:08:43.279 --> 00:08:50.080 Obviously,<00:08:43.525> um,<00:08:43.771> one<00:08:44.017> of<00:08:44.263> the<00:08:44.509> things<00:08:44.755> that<00:08:45.001> came<00:08:45.247> out,<00:08:45.493> um,<00:08:45.739> over<00:08:45.985> the<00:08:46.231> past<00:08:46.477> few<00:08:46.722> months<00:08:46.968> that<00:08:47.214> obviously<00:08:47.460> got<00:08:47.706> a<00:08:47.952> lot<00:08:48.198> of<00:08:48.444> traction,<00:08:48.690> I<00:08:48.936> think<00:08:49.182> it's<00:08:49.428> probably<00:08:49.674> like<00:08:49.920> the 00:08:50.080 --> 00:08:58.480 biggest<00:08:50.457> GitHub<00:08:50.834> project<00:08:51.211> in<00:08:51.589> history<00:08:51.966> is<00:08:52.343> OpenClaw,<00:08:52.720> right,<00:08:53.097> and<00:08:53.474> everything<00:08:53.851> that,<00:08:54.229> um,<00:08:54.606> is<00:08:54.983> happening<00:08:55.360> around<00:08:55.737> there,<00:08:56.114> obviously<00:08:56.491> that's<00:08:56.869> open<00:08:57.246> source<00:08:57.623> security<00:08:58.000> is 00:08:58.480 --> 00:09:06.959 a<00:08:58.720> huge<00:08:58.959> issue<00:08:59.680> around,<00:09:00.560> uh,<00:09:00.800> projects<00:09:01.200> like<00:09:01.440> that.<00:09:02.320> What<00:09:02.560> can<00:09:02.800> companies<00:09:03.600> and<00:09:04.080> entrepreneurs<00:09:04.800> and<00:09:05.120> founders<00:09:05.440> and<00:09:05.680> CEOs<00:09:06.160> do<00:09:06.399> to<00:09:06.640> protect 00:09:06.959 --> 00:09:16.240 themselves<00:09:07.509> against<00:09:08.059> this?<00:09:08.609> So<00:09:09.159> look,<00:09:09.709> OpenClaw<00:09:10.259> was<00:09:10.809> just<00:09:11.360> an<00:09:11.910> outlier<00:09:12.460> anomaly<00:09:13.010> because<00:09:13.560> of<00:09:14.110> just<00:09:14.660> the<00:09:15.210> overnight<00:09:15.760> claim 00:09:16.240 --> 00:09:24.959 to<00:09:16.560> fame.<00:09:18.320> I<00:09:18.640> actually<00:09:18.880> think<00:09:19.120> OpenClaw<00:09:20.720> was<00:09:21.040> incredibly<00:09:21.680> powerful<00:09:22.080> to<00:09:22.399> show<00:09:22.560> us<00:09:23.680> what<00:09:23.920> is<00:09:24.160> possible<00:09:24.560> when 00:09:24.959 --> 00:09:28.560 agents<00:09:25.360> and<00:09:25.680> AI<00:09:26.720> starts<00:09:27.040> to<00:09:27.279> outsmart<00:09:27.920> individuals. 00:09:28.560 --> 00:09:35.200 If<00:09:28.640> you<00:09:28.800> if<00:09:28.959> you've<00:09:29.120> seen<00:09:29.279> some<00:09:29.440> of<00:09:29.519> the<00:09:29.760> stories,<00:09:30.800> basically<00:09:31.200> the<00:09:31.440> way<00:09:31.519> it<00:09:31.680> was<00:09:31.839> built<00:09:32.080> to<00:09:32.240> reason<00:09:32.480> with<00:09:32.720> each<00:09:32.880> other<00:09:33.200> was<00:09:33.279> kind<00:09:33.440> of<00:09:33.519> taking<00:09:33.839> over<00:09:34.560> from<00:09:34.880> others. 00:09:35.200 --> 00:09:38.800 But<00:09:35.360> like<00:09:35.519> what<00:09:35.680> it<00:09:35.920> did<00:09:36.080> showcase<00:09:37.279> is<00:09:37.519> the<00:09:37.760> power<00:09:37.920> of<00:09:38.160> open<00:09:38.399> source. 00:09:38.800 --> 00:09:44.240 If<00:09:39.040> you<00:09:39.120> put<00:09:39.279> something<00:09:39.600> good<00:09:39.839> out<00:09:40.000> there<00:09:40.720> overnight,<00:09:41.920> millions<00:09:42.320> of<00:09:42.560> people<00:09:42.800> will<00:09:43.120> start<00:09:43.360> using<00:09:43.680> it. 00:09:44.240 --> 00:09:52.640 Now<00:09:44.480> the<00:09:44.640> question<00:09:44.880> from<00:09:45.360> software<00:09:45.760> supply<00:09:46.080> chain<00:09:46.399> perspective<00:09:47.360> and<00:09:47.600> it<00:09:47.839> did<00:09:48.000> actually<00:09:48.240> come<00:09:48.399> out<00:09:48.560> in<00:09:48.720> the<00:09:48.880> next<00:09:49.519> subsequent<00:09:50.080> few<00:09:50.320> days<00:09:51.279> is<00:09:52.240> the 00:09:52.640 --> 00:09:56.160 security<00:09:52.959> of<00:09:53.200> that<00:09:53.440> software<00:09:53.760> as<00:09:53.920> it<00:09:54.080> was<00:09:54.240> built<00:09:54.480> was<00:09:54.720> terrible. 00:09:56.160 --> 00:10:02.640 But<00:09:56.240> it<00:09:56.480> was<00:09:56.560> never<00:09:56.800> built<00:09:57.120> with<00:09:57.279> somebody<00:09:57.760> thinking<00:09:58.160> that<00:09:58.320> they<00:09:58.480> would<00:09:58.640> have<00:09:58.800> such<00:09:59.120> overnight<00:09:59.680> success<00:10:00.560> and<00:10:00.880> the<00:10:01.120> implications<00:10:01.760> of<00:10:01.920> it<00:10:02.080> going<00:10:02.320> wrong. 00:10:02.640 --> 00:10:05.519 And<00:10:02.800> sadly<00:10:03.200> that<00:10:03.440> is<00:10:03.519> the<00:10:03.760> state<00:10:04.000> of<00:10:04.240> software. 00:10:05.519 --> 00:10:09.839 Security<00:10:06.000> has<00:10:06.240> always<00:10:06.399> been<00:10:06.560> an<00:10:06.800> afterthought<00:10:08.080> in<00:10:08.399> any<00:10:08.720> company<00:10:08.959> on<00:10:09.200> the<00:10:09.360> planet,<00:10:09.600> right? 00:10:09.839 --> 00:10:17.200 You<00:10:10.000> ship,<00:10:11.200> you<00:10:11.440> become<00:10:11.760> successful,<00:10:12.640> then<00:10:12.880> your<00:10:13.040> customers<00:10:13.440> ask<00:10:13.680> you<00:10:13.760> about<00:10:14.000> security,<00:10:14.399> then<00:10:14.560> you<00:10:14.720> think<00:10:14.880> about<00:10:15.120> security,<00:10:15.839> and<00:10:16.000> even<00:10:16.240> then<00:10:16.399> you're<00:10:16.640> first<00:10:16.880> thinking 00:10:17.200 --> 00:10:21.600 about<00:10:17.360> compliance.<00:10:17.920> Like,<00:10:18.079> what<00:10:18.160> do<00:10:18.240> I<00:10:18.399> need<00:10:18.480> to<00:10:18.640> do<00:10:18.720> to<00:10:18.880> check<00:10:19.040> a<00:10:19.279> box<00:10:19.680> so<00:10:19.920> I<00:10:20.079> can<00:10:20.240> make<00:10:20.320> a<00:10:20.560> sale? 00:10:21.600 --> 00:10:25.279 And<00:10:21.760> then<00:10:21.920> your<00:10:22.160> business<00:10:22.399> becomes<00:10:22.800> consequential<00:10:23.440> enough<00:10:23.680> that<00:10:23.920> you<00:10:24.079> think<00:10:24.240> about<00:10:24.399> real<00:10:24.800> security. 00:10:25.279 --> 00:10:28.240 That's<00:10:25.519> just<00:10:25.680> the<00:10:25.920> maturity<00:10:26.320> curve<00:10:26.640> we<00:10:26.800> live<00:10:26.959> in. 00:10:28.240 --> 00:10:37.040 Except<00:10:28.640> today<00:10:30.079> none<00:10:30.320> of<00:10:30.480> that<00:10:30.640> will<00:10:30.880> work<00:10:31.200> because<00:10:31.519> no<00:10:31.600> matter<00:10:31.920> who<00:10:32.079> you<00:10:32.320> are,<00:10:32.640> your<00:10:33.040> software<00:10:34.000> will<00:10:34.320> be<00:10:34.480> compromised<00:10:35.120> if<00:10:35.440> it<00:10:35.600> is<00:10:35.680> of<00:10:36.000> consequence<00:10:36.800> to 00:10:37.040 --> 00:10:43.279 the<00:10:37.279> industry,<00:10:38.480> uh,<00:10:38.640> if<00:10:38.800> you<00:10:38.959> don't<00:10:39.120> get<00:10:39.360> security<00:10:39.920> embedded<00:10:40.560> into<00:10:40.880> your<00:10:41.120> software<00:10:41.519> development<00:10:41.920> workflow. 00:10:43.279 --> 00:10:51.120 And<00:10:43.760> with<00:10:44.079> OpenClaw<00:10:44.560> it<00:10:44.640> was<00:10:44.800> the<00:10:44.880> same<00:10:45.040> thing,<00:10:45.200> like<00:10:45.519> thankfully<00:10:46.000> it<00:10:46.160> got<00:10:46.959> inherited<00:10:47.440> by<00:10:47.600> a<00:10:47.760> large<00:10:48.000> frontier<00:10:48.480> lab<00:10:48.720> and<00:10:49.360> now<00:10:49.600> they're<00:10:49.839> figuring<00:10:50.000> out<00:10:50.160> what<00:10:50.320> to<00:10:50.480> do<00:10:50.880> with 00:10:51.120 --> 00:10:56.800 the<00:10:51.279> rest<00:10:51.440> of<00:10:51.519> that<00:10:51.760> software.<00:10:52.240> And<00:10:52.320> I<00:10:52.480> think<00:10:52.640> there<00:10:52.880> was<00:10:53.040> a<00:10:53.200> lot<00:10:53.360> of<00:10:53.440> warnings<00:10:53.839> very<00:10:54.160> quickly,<00:10:54.480> people<00:10:54.720> saying<00:10:54.880> like,<00:10:55.200> listen<00:10:55.440> guys,<00:10:55.680> this<00:10:55.839> is<00:10:55.920> not<00:10:56.079> for<00:10:56.240> hobbyists,<00:10:56.640> like 00:10:56.800 --> 00:11:03.920 if<00:10:57.035> you<00:10:57.269> don't<00:10:57.504> know<00:10:57.739> what<00:10:57.973> you're<00:10:58.208> doing,<00:10:58.442> don't<00:10:58.677> turn<00:10:58.912> this<00:10:59.146> on.<00:10:59.381> But<00:10:59.616> it<00:10:59.850> did<00:11:00.085> explain<00:11:00.320> the<00:11:00.554> power<00:11:00.789> of<00:11:01.023> open<00:11:01.258> source<00:11:01.493> and<00:11:01.727> how<00:11:01.962> fast<00:11:02.197> it,<00:11:02.431> uh,<00:11:02.666> uh,<00:11:02.900> it<00:11:03.135> expands<00:11:03.370> in<00:11:03.604> the<00:11:03.839> fact 00:11:03.920 --> 00:11:10.399 that<00:11:04.160> today<00:11:04.959> we<00:11:05.279> need<00:11:05.519> security<00:11:06.000> to<00:11:06.240> be<00:11:06.399> your<00:11:06.720> default<00:11:07.920> way<00:11:08.160> of<00:11:08.399> shipping<00:11:08.720> code<00:11:09.279> and<00:11:09.519> not<00:11:09.680> an<00:11:09.839> afterthought. 00:11:10.399 --> 00:11:18.079 And<00:11:10.480> now<00:11:10.640> we<00:11:10.800> can<00:11:10.880> kind<00:11:11.040> of<00:11:11.120> go<00:11:11.279> down,<00:11:11.760> I<00:11:11.920> I<00:11:12.240> have<00:11:12.320> a<00:11:12.480> whole<00:11:12.720> spiel<00:11:13.040> on<00:11:13.920> why<00:11:14.160> have<00:11:14.320> we<00:11:14.480> as<00:11:14.640> an<00:11:14.800> industry<00:11:15.120> failed<00:11:15.440> for<00:11:15.600> the<00:11:15.760> last<00:11:15.920> 30<00:11:16.160> years<00:11:16.320> to<00:11:16.560> ship<00:11:16.800> secure<00:11:17.279> software. 00:11:18.079 --> 00:11:20.640 A<00:11:18.240> big<00:11:18.480> part<00:11:18.640> of<00:11:18.720> that<00:11:18.959> has<00:11:19.120> certainly<00:11:19.440> been<00:11:19.600> it's<00:11:19.920> been<00:11:20.000> hard, 00:11:20.640 --> 00:11:28.240 expensive,<00:11:21.440> and<00:11:21.680> trying<00:11:21.920> to<00:11:22.079> get<00:11:22.320> thousands<00:11:22.720> and<00:11:22.959> thousands<00:11:23.279> of<00:11:23.440> developers<00:11:23.920> to<00:11:24.160> consistently<00:11:24.959> pay<00:11:25.120> attention<00:11:25.440> to<00:11:25.680> do<00:11:25.920> things<00:11:26.160> right<00:11:26.959> is<00:11:27.360> hard. 00:11:28.240 --> 00:11:33.040 But<00:11:28.399> with<00:11:28.640> agents,<00:11:28.959> you<00:11:29.200> have<00:11:29.279> the<00:11:29.519> opportunity<00:11:29.920> to<00:11:30.240> give<00:11:30.399> them<00:11:30.560> the<00:11:30.800> right<00:11:30.959> coaching,<00:11:31.440> give<00:11:31.519> them<00:11:31.680> the<00:11:31.839> right<00:11:32.000> instructions. 00:11:33.040 --> 00:11:41.200 And<00:11:33.200> if<00:11:33.360> you<00:11:33.519> do<00:11:33.680> embed<00:11:34.640> the<00:11:34.959> workflow<00:11:35.519> of<00:11:35.839> security<00:11:36.240> and<00:11:36.480> the<00:11:36.640> security<00:11:37.040> intelligence<00:11:38.399> as<00:11:38.959> we<00:11:39.120> do<00:11:39.279> at<00:11:39.440> Endor<00:11:39.839> and<00:11:40.000> we<00:11:40.160> build<00:11:40.399> products<00:11:40.640> to<00:11:40.880> do<00:11:41.040> for 00:11:41.200 --> 00:11:49.600 our<00:11:41.440> customers,<00:11:42.800> you<00:11:43.120> can<00:11:43.279> actually<00:11:43.519> get<00:11:43.839> security<00:11:44.160> to<00:11:44.320> be<00:11:44.480> more<00:11:44.720> consistent<00:11:45.120> as<00:11:45.360> an<00:11:45.600> outcome<00:11:46.079> in<00:11:46.320> your<00:11:46.560> code<00:11:46.800> generation<00:11:47.279> workflow. 00:11:49.600 --> 00:11:56.160 So,<00:11:50.480> let<00:11:50.720> me<00:11:50.800> ask<00:11:50.959> you<00:11:51.040> this<00:11:51.279> about<00:11:51.760> uh<00:11:51.839> because<00:11:52.079> a<00:11:52.240> lot<00:11:52.320> of<00:11:52.399> people<00:11:52.560> are<00:11:52.720> thinking,<00:11:52.959> well,<00:11:53.120> I'm<00:11:53.279> going<00:11:53.360> to<00:11:53.440> need<00:11:53.519> to<00:11:53.680> have<00:11:53.839> everything<00:11:54.079> on<00:11:54.399> prem,<00:11:54.800> nothing<00:11:55.120> in<00:11:55.360> the<00:11:55.519> cloud, 00:11:56.160 --> 00:11:59.760 have<00:11:56.320> my<00:11:56.560> own<00:11:57.360> uh<00:11:58.079> you<00:11:58.240> know,<00:11:58.399> data<00:11:58.720> center<00:11:58.959> within<00:11:59.279> the<00:11:59.440> company. 00:11:59.760 --> 00:12:04.320 What<00:11:59.920> what's<00:12:00.240> your<00:12:01.519> take<00:12:01.760> for<00:12:02.000> those<00:12:02.160> who<00:12:02.320> are<00:12:02.480> kind<00:12:02.560> of<00:12:02.640> thinking<00:12:02.880> that<00:12:03.200> route? 00:12:04.320 --> 00:12:07.519 It's<00:12:04.480> all<00:12:04.720> false<00:12:05.040> sense<00:12:05.279> of<00:12:05.440> security.<00:12:06.000> Yeah,<00:12:06.240> it's<00:12:06.480> not<00:12:06.639> going<00:12:06.720> to<00:12:06.880> work. 00:12:07.519 --> 00:12:13.680 Uh<00:12:07.839> look,<00:12:08.000> when<00:12:08.160> I<00:12:08.320> was<00:12:08.399> building<00:12:08.639> Redlock<00:12:09.120> in<00:12:09.360> 2015,<00:12:10.399> the<00:12:10.639> whole<00:12:10.800> idea<00:12:11.120> was<00:12:11.920> you're<00:12:12.240> going<00:12:12.320> to<00:12:12.399> have<00:12:12.480> to<00:12:12.639> secure<00:12:12.959> the<00:12:13.120> public<00:12:13.360> cloud. 00:12:13.680 --> 00:12:21.279 And<00:12:13.941> a<00:12:14.201> lot<00:12:14.462> of<00:12:14.723> people<00:12:14.984> were<00:12:15.244> like,<00:12:15.505> well,<00:12:15.766> I'm<00:12:16.027> afraid<00:12:16.287> of<00:12:16.548> public<00:12:16.809> cloud,<00:12:17.070> so<00:12:17.330> why<00:12:17.591> don't<00:12:17.852> I<00:12:18.113> do<00:12:18.373> private<00:12:18.634> cloud<00:12:18.895> on<00:12:19.156> VMware,<00:12:19.416> bare<00:12:19.677> metal,<00:12:19.938> create<00:12:20.199> my<00:12:20.459> own<00:12:20.720> OpenShift 00:12:21.279 --> 00:12:24.800 clusters.<00:12:22.000> And<00:12:22.160> you<00:12:22.320> know<00:12:22.399> what?<00:12:23.440> They<00:12:23.680> all<00:12:23.920> failed. 00:12:24.800 --> 00:12:33.279 And<00:12:24.959> I<00:12:25.200> would<00:12:25.360> say<00:12:26.160> the<00:12:26.399> the<00:12:26.880> value<00:12:27.200> prop,<00:12:27.600> the<00:12:27.839> business<00:12:28.160> benefit<00:12:28.480> of<00:12:28.720> moving<00:12:28.959> to<00:12:29.120> public<00:12:29.440> cloud<00:12:29.760> was<00:12:30.000> so<00:12:30.320> significant<00:12:31.920> that<00:12:32.320> you<00:12:32.639> actually<00:12:33.040> put 00:12:33.279 --> 00:12:37.279 your<00:12:33.760> company<00:12:34.079> at<00:12:34.320> a<00:12:34.480> competitive<00:12:35.040> disadvantage<00:12:35.760> by<00:12:36.079> not<00:12:36.320> doing<00:12:36.560> that. 00:12:37.279 --> 00:12:42.480 And<00:12:37.440> I<00:12:37.680> would<00:12:37.760> say<00:12:37.839> the<00:12:38.079> stakes<00:12:38.399> with<00:12:38.639> AI<00:12:39.519> are<00:12:39.839> 10<00:12:40.079> times<00:12:40.399> higher,<00:12:41.040> 100<00:12:41.440> times<00:12:41.680> higher. 00:12:42.480 --> 00:12:46.240 The<00:12:42.800> consequence<00:12:43.200> of<00:12:43.440> not<00:12:43.600> getting<00:12:43.760> it<00:12:44.000> right<00:12:44.240> is<00:12:44.480> that<00:12:44.720> much<00:12:44.800> higher<00:12:45.040> and<00:12:45.200> you<00:12:45.360> just<00:12:45.519> can't<00:12:45.680> slow<00:12:45.839> it<00:12:46.000> down. 00:12:46.240 --> 00:12:54.720 So<00:12:46.320> I<00:12:46.480> talked<00:12:46.639> to<00:12:46.720> a<00:12:46.880> lot<00:12:46.959> of<00:12:47.040> security<00:12:47.360> teams,<00:12:48.079> a<00:12:48.240> lot<00:12:48.320> of<00:12:48.480> leaders<00:12:48.880> where<00:12:50.000> you<00:12:50.160> know<00:12:51.360> honestly<00:12:51.760> it<00:12:52.000> was<00:12:52.079> less<00:12:52.320> than<00:12:52.480> a<00:12:52.720> year<00:12:53.200> before<00:12:53.680> every<00:12:54.160> company<00:12:54.480> on 00:12:54.720 --> 00:13:02.240 the<00:12:54.800> planet<00:12:55.120> was<00:12:55.440> perfectly<00:12:56.000> comfortable<00:12:56.880> with<00:12:57.200> shipping<00:12:57.519> their<00:12:57.760> sensitive<00:12:58.240> data<00:12:59.200> to<00:12:59.519> these<00:12:59.760> frontier<00:13:00.240> models<00:13:00.639> with<00:13:00.880> the<00:13:01.040> right<00:13:01.279> contracts<00:13:01.760> in<00:13:02.000> place. 00:13:02.240 --> 00:13:06.240 Right?<00:13:02.480> So<00:13:02.639> we<00:13:02.800> kind<00:13:02.959> of<00:13:03.040> got<00:13:03.200> through<00:13:03.440> it<00:13:03.600> way<00:13:03.920> faster<00:13:04.240> than<00:13:04.399> we<00:13:04.560> did<00:13:04.720> with<00:13:04.959> cloud. 00:13:06.240 --> 00:13:12.959 The<00:13:06.480> people<00:13:06.880> talking<00:13:07.200> about<00:13:07.360> like<00:13:07.600> let<00:13:07.760> me<00:13:07.920> have<00:13:08.000> my<00:13:08.240> own<00:13:08.399> models<00:13:08.720> and<00:13:08.959> host<00:13:09.200> my<00:13:09.360> own<00:13:09.440> models<00:13:09.839> quickly<00:13:10.160> realize<00:13:10.560> you<00:13:10.800> can't<00:13:10.959> be<00:13:11.120> competitive<00:13:12.079> because<00:13:12.320> the<00:13:12.560> frontier 00:13:12.959 --> 00:13:15.839 models<00:13:13.279> are<00:13:13.440> getting<00:13:13.760> so<00:13:14.000> much<00:13:14.240> better<00:13:14.560> so<00:13:14.800> much<00:13:14.959> faster. 00:13:15.839 --> 00:13:20.639 Now<00:13:16.079> granted<00:13:16.480> there's<00:13:16.720> a<00:13:16.800> whole<00:13:17.040> cost<00:13:17.680> issue<00:13:18.160> and<00:13:18.880> token<00:13:19.279> maxing<00:13:19.680> and<00:13:19.839> other<00:13:20.000> things<00:13:20.160> that's<00:13:20.399> coming. 00:13:20.639 --> 00:13:23.120 We'll<00:13:20.880> we'll<00:13:21.040> I'm<00:13:21.200> sure<00:13:21.440> touch<00:13:21.600> on<00:13:21.680> that<00:13:21.839> in<00:13:22.000> a<00:13:22.160> moment. 00:13:23.120 --> 00:13:29.519 But<00:13:23.279> like<00:13:23.600> my<00:13:23.920> counsel<00:13:25.040> to<00:13:25.519> adopters<00:13:26.079> of<00:13:26.320> of<00:13:26.800> AI<00:13:28.000> is<00:13:28.320> don't<00:13:28.560> slow<00:13:28.880> it<00:13:29.120> down. 00:13:29.519 --> 00:13:37.760 Don't<00:13:29.760> look<00:13:29.920> for<00:13:30.160> alternatives.<00:13:30.880> Just<00:13:31.120> figure<00:13:31.279> out<00:13:31.440> how<00:13:31.600> to<00:13:32.079> take<00:13:32.240> the<00:13:32.399> bull<00:13:32.639> by<00:13:32.800> its<00:13:32.959> horn<00:13:33.600> and<00:13:34.000> instrument<00:13:34.480> and<00:13:34.720> insert<00:13:35.279> security<00:13:35.760> intelligence<00:13:36.720> and<00:13:37.040> safety<00:13:37.440> guard 00:13:37.760 --> 00:13:42.399 rails<00:13:38.560> into<00:13:39.040> the<00:13:39.279> adoption<00:13:40.160> versus<00:13:40.560> trying<00:13:40.720> to<00:13:40.880> hold<00:13:41.120> back<00:13:41.360> the<00:13:41.600> adoption. 00:13:42.399 --> 00:13:49.760 You<00:13:43.040> are<00:13:43.200> not<00:13:43.440> making<00:13:43.600> a<00:13:43.839> good<00:13:44.079> career<00:13:44.399> move<00:13:45.040> if<00:13:45.279> you're<00:13:45.440> trying<00:13:45.519> to<00:13:45.680> hold<00:13:45.920> back<00:13:46.320> adoption<00:13:46.800> of<00:13:46.880> AI<00:13:47.200> in<00:13:47.360> your<00:13:47.519> company<00:13:47.839> right<00:13:48.079> now. 00:13:49.760 --> 00:13:54.639 How<00:13:49.920> much<00:13:50.079> code<00:13:50.320> do<00:13:50.480> you<00:13:50.560> think<00:13:50.720> has<00:13:50.880> been<00:13:50.959> written<00:13:51.200> that<00:13:51.440> needs<00:13:51.600> to<00:13:51.680> be<00:13:51.839> rewritten 00:13:54.639 --> 00:14:02.320 over<00:13:54.922> the<00:13:55.205> entirety<00:13:55.488> of<00:13:55.771> time?<00:13:56.055> You're<00:13:56.338> talking<00:13:56.621> specifically<00:13:56.904> with<00:13:57.187> AI<00:13:57.470> or<00:13:57.753> <00:13:58.036> Yeah,<00:13:58.320> just<00:13:58.603> I<00:13:58.886> don't<00:13:59.169> Well,<00:13:59.452> I<00:13:59.735> mean<00:14:00.018> let's<00:14:00.301> take<00:14:00.584> like<00:14:00.868> uh<00:14:01.151> for<00:14:01.434> example<00:14:01.717> like<00:14:02.000> the 00:14:02.320 --> 00:14:09.760 government.<00:14:02.569> <00:14:02.817> I<00:14:03.066> just<00:14:03.314> want<00:14:03.563> to<00:14:03.811> get<00:14:04.060> like<00:14:04.308> a<00:14:04.557> sense<00:14:04.805> for<00:14:05.054> like<00:14:05.302> the<00:14:05.551> scale<00:14:05.800> and<00:14:06.048> get<00:14:06.297> people<00:14:06.545> to<00:14:06.794> understand<00:14:07.042> the<00:14:07.291> scale<00:14:07.539> like<00:14:07.788> uh<00:14:08.036> how<00:14:08.285> much<00:14:08.533> software<00:14:08.782> has<00:14:09.030> been<00:14:09.279> rewritten 00:14:09.760 --> 00:14:18.000 that<00:14:10.046> now<00:14:10.331> given<00:14:10.617> the<00:14:10.903> potentials<00:14:11.189> of<00:14:11.474> AI<00:14:11.760> to<00:14:12.046> kind<00:14:12.331> of<00:14:12.617> uh<00:14:12.903> make<00:14:13.189> it<00:14:13.474> much<00:14:13.760> easier<00:14:14.046> to<00:14:14.331> hack<00:14:14.617> <00:14:14.903> but<00:14:15.189> now<00:14:15.474> also<00:14:15.760> given<00:14:16.046> the<00:14:16.331> potentials<00:14:16.617> of<00:14:16.903> AI<00:14:17.189> to<00:14:17.474> make<00:14:17.760> this 00:14:18.000 --> 00:14:24.480 software<00:14:18.399> better.<00:14:18.720> How<00:14:18.880> much<00:14:19.040> of<00:14:19.120> it<00:14:19.360> needs<00:14:19.600> to<00:14:19.680> be<00:14:19.920> rewritten<00:14:20.480> from<00:14:20.639> the<00:14:20.880> ground<00:14:21.040> up? 00:14:24.480 --> 00:14:29.040 <00:14:24.739> Hundreds<00:14:24.998> of<00:14:25.256> billions<00:14:25.515> of<00:14:25.774> lines<00:14:26.033> of<00:14:26.292> code,<00:14:26.551> maybe<00:14:26.809> even<00:14:27.068> as<00:14:27.327> much<00:14:27.586> as<00:14:27.845> trillions<00:14:28.104> of<00:14:28.362> lines<00:14:28.621> of<00:14:28.880> code. 00:14:29.040 --> 00:14:36.000 If<00:14:29.199> you<00:14:29.279> look<00:14:29.360> at<00:14:29.440> the<00:14:29.600> entirety<00:14:29.920> of<00:14:30.079> the<00:14:30.240> software<00:14:30.720> state<00:14:30.959> we<00:14:31.279> live<00:14:31.440> on<00:14:32.079> and<00:14:32.320> just<00:14:32.560> how<00:14:33.279> horrible<00:14:34.320> the<00:14:34.639> security<00:14:34.959> state<00:14:35.199> of<00:14:35.360> affairs<00:14:35.680> is. 00:14:36.000 --> 00:14:44.320 Just<00:14:36.160> to<00:14:36.320> give<00:14:36.560> the<00:14:36.800> listeners<00:14:37.199> an<00:14:37.600> you<00:14:37.760> know<00:14:38.480> orders<00:14:38.880> of<00:14:39.120> magnitude<00:14:40.399> issue<00:14:40.880> like<00:14:41.199> a<00:14:42.079> average<00:14:42.560> company<00:14:43.279> prethos<00:14:44.160> like 00:14:44.320 --> 00:14:52.800 models<00:14:44.639> that<00:14:44.800> are<00:14:44.959> supposed<00:14:45.120> to<00:14:45.279> get<00:14:45.519> even<00:14:45.760> better.<00:14:47.120> An<00:14:47.440> average<00:14:48.079> financial<00:14:48.639> organization,<00:14:49.600> financial<00:14:50.079> services<00:14:50.720> bank<00:14:51.680> is<00:14:51.920> sitting<00:14:52.160> on<00:14:52.399> about 00:14:52.800 --> 00:15:01.120 5<00:14:53.199> to<00:14:53.600> 10<00:14:53.920> million<00:14:55.199> known<00:14:55.760> security<00:14:57.360> security<00:14:57.839> vulnerabilities<00:14:58.560> in<00:14:58.800> their<00:14:59.199> existing<00:14:59.760> software. 00:15:01.120 --> 00:15:06.480 And<00:15:01.360> this<00:15:01.519> is<00:15:01.680> just<00:15:01.839> in<00:15:02.000> the<00:15:02.160> software<00:15:02.560> they<00:15:02.880> built,<00:15:03.279> not<00:15:03.440> the<00:15:03.680> software<00:15:04.000> they've<00:15:04.399> licensed<00:15:04.880> from<00:15:05.040> third<00:15:05.360> parties. 00:15:06.480 --> 00:15:15.600 Now<00:15:06.720> I<00:15:06.959> say<00:15:07.199> that<00:15:07.360> with<00:15:07.680> air<00:15:07.839> quotes<00:15:08.320> because<00:15:09.120> part<00:15:09.360> of<00:15:09.440> the<00:15:09.680> problem<00:15:09.839> in<00:15:10.000> the<00:15:10.160> industry<00:15:10.959> that<00:15:11.279> certainly<00:15:11.600> we've<00:15:11.839> set<00:15:12.000> out<00:15:12.160> to<00:15:12.320> solve<00:15:12.560> with<00:15:12.720> Endor<00:15:13.120> Labs<00:15:14.160> is<00:15:15.279> we 00:15:15.600 --> 00:15:24.800 know<00:15:15.760> the<00:15:16.240> all<00:15:16.480> not<00:15:16.800> all<00:15:17.199> 5<00:15:17.440> to<00:15:17.680> 10<00:15:18.000> million<00:15:18.880> findings<00:15:19.519> are<00:15:20.560> necessarily<00:15:21.120> going<00:15:21.279> to<00:15:21.360> affect<00:15:21.680> you<00:15:21.839> or<00:15:22.160> be<00:15:22.320> the<00:15:22.560> things<00:15:22.720> that<00:15:22.880> put<00:15:23.040> you<00:15:23.199> in<00:15:23.360> the<00:15:23.519> news. 00:15:24.800 --> 00:15:32.959 However,<00:15:25.440> trying<00:15:25.680> to<00:15:26.560> go<00:15:26.800> through<00:15:27.040> and<00:15:27.360> triage<00:15:27.839> those<00:15:28.160> 5<00:15:28.399> to<00:15:28.560> 10<00:15:28.800> million<00:15:29.120> to<00:15:29.360> figure<00:15:29.519> out<00:15:29.760> which<00:15:31.199> call<00:15:31.360> it<00:15:31.680> half<00:15:31.839> a<00:15:32.079> million<00:15:32.560> which<00:15:32.800> is 00:15:32.959 --> 00:15:42.240 also<00:15:33.199> still<00:15:33.440> a<00:15:33.600> really<00:15:33.839> big<00:15:34.079> number<00:15:34.959> are<00:15:35.199> real<00:15:36.560> is<00:15:36.800> going<00:15:36.959> to<00:15:37.120> take<00:15:37.360> you<00:15:37.600> 10<00:15:37.920> years<00:15:38.160> to<00:15:38.399> do<00:15:39.199> if<00:15:39.440> not<00:15:39.600> more<00:15:40.160> with<00:15:40.399> your<00:15:41.279> existing 00:15:42.240 --> 00:15:51.360 workflows<00:15:42.880> and<00:15:43.120> technology.<00:15:44.240> So<00:15:44.480> the<00:15:44.720> first<00:15:44.880> things<00:15:45.279> first<00:15:45.920> is<00:15:46.160> you<00:15:46.399> have<00:15:46.480> to<00:15:46.639> invest<00:15:47.040> in<00:15:48.160> agentic<00:15:48.800> technology<00:15:49.360> to<00:15:49.600> fight<00:15:51.040> the 00:15:51.360 --> 00:15:59.600 agentic<00:15:52.160> adversaries<00:15:53.279> who<00:15:53.519> have<00:15:53.759> gotten<00:15:54.000> to<00:15:54.160> this<00:15:54.480> faster<00:15:54.800> than<00:15:55.040> you<00:15:55.279> have<00:15:56.639> and<00:15:56.959> really<00:15:57.199> go<00:15:57.519> try<00:15:57.680> to<00:15:57.759> figure<00:15:57.920> out<00:15:58.160> how<00:15:58.320> do<00:15:58.480> I<00:15:58.800> go<00:15:59.120> suss<00:15:59.360> out 00:15:59.600 --> 00:16:07.920 what's<00:15:59.920> real<00:16:00.240> what's<00:16:00.560> not<00:16:01.040> which<00:16:01.199> of<00:16:01.360> these<00:16:01.519> problems<00:16:01.839> are<00:16:02.000> real<00:16:03.120> and<00:16:03.440> we<00:16:03.680> see<00:16:03.920> data<00:16:04.240> that<00:16:04.399> shows<00:16:04.720> our<00:16:04.959> customers<00:16:05.360> when<00:16:05.519> you<00:16:05.839> apply<00:16:06.240> technology<00:16:07.680> it 00:16:07.920 --> 00:16:16.560 turns<00:16:08.160> out<00:16:08.320> only<00:16:08.560> about<00:16:08.800> 2<00:16:08.959> to<00:16:09.199> 5%<00:16:09.759> of<00:16:09.920> the<00:16:10.160> things<00:16:10.320> that<00:16:10.480> are<00:16:10.639> theoretically<00:16:11.279> being<00:16:11.440> reported<00:16:11.759> by<00:16:12.000> your<00:16:12.480> legacy<00:16:12.959> software<00:16:13.360> security<00:16:13.759> tools<00:16:14.160> are<00:16:14.480> real,<00:16:16.240> but 00:16:16.560 --> 00:16:24.320 then<00:16:16.880> you<00:16:17.120> get<00:16:17.279> stuck<00:16:17.519> at<00:16:17.680> the<00:16:17.759> altar<00:16:18.320> because<00:16:19.279> I<00:16:19.600> know<00:16:19.839> what<00:16:20.880> I<00:16:21.199> need<00:16:21.279> to<00:16:21.440> fix,<00:16:21.920> but<00:16:22.240> it's<00:16:22.480> really<00:16:22.800> freaking<00:16:23.199> hard<00:16:23.360> to<00:16:23.600> fix<00:16:23.839> it. 00:16:24.320 --> 00:16:32.480 Changing<00:16:24.800> code<00:16:24.959> that<00:16:25.199> was<00:16:25.360> written<00:16:25.600> 9<00:16:25.839> years<00:16:25.920> ago,<00:16:26.240> 8<00:16:26.399> years<00:16:26.560> ago,<00:16:27.199> testing<00:16:27.519> it,<00:16:27.839> recertifying<00:16:28.480> it,<00:16:28.720> making<00:16:28.880> sure<00:16:29.040> it<00:16:29.199> doesn't<00:16:29.360> break<00:16:29.680> something<00:16:30.000> else<00:16:30.720> inconsequentially 00:16:32.480 --> 00:16:38.000 is<00:16:32.800> hard.<00:16:33.759> And<00:16:33.920> so<00:16:34.160> now<00:16:34.320> the<00:16:34.560> next<00:16:34.720> question<00:16:34.880> becomes,<00:16:35.279> if<00:16:35.440> I<00:16:35.680> figured<00:16:35.920> out<00:16:36.079> what<00:16:36.240> to<00:16:36.480> fix,<00:16:37.040> how<00:16:37.199> do<00:16:37.279> I<00:16:37.519> automate? 00:16:38.000 --> 00:16:40.720 How<00:16:38.079> do<00:16:38.160> I<00:16:38.320> use<00:16:38.480> the<00:16:38.639> agents<00:16:39.440> to<00:16:39.759> reason<00:16:40.079> with<00:16:40.240> my<00:16:40.399> code? 00:16:40.720 --> 00:16:48.639 Give<00:16:40.959> them<00:16:41.120> the<00:16:41.360> right<00:16:41.600> context<00:16:42.560> so<00:16:42.800> they<00:16:43.040> just<00:16:43.199> know<00:16:43.360> how<00:16:43.519> to<00:16:43.680> fix<00:16:43.839> it<00:16:44.079> because<00:16:44.320> I<00:16:44.560> cannot<00:16:44.880> expect<00:16:45.199> humans<00:16:45.600> to<00:16:45.839> fix<00:16:46.480> even<00:16:46.880> half<00:16:47.120> a<00:16:47.279> million<00:16:47.680> problems. 00:16:48.639 --> 00:16:53.680 And<00:16:48.800> so<00:16:48.959> like<00:16:49.199> we<00:16:49.360> have<00:16:49.519> years<00:16:49.839> and<00:16:50.079> years<00:16:50.320> of<00:16:51.040> debt<00:16:51.440> that<00:16:51.680> we<00:16:51.920> have<00:16:52.079> to<00:16:52.880> address. 00:16:53.680 --> 00:17:02.000 But<00:16:53.839> at<00:16:54.079> the<00:16:54.240> same<00:16:54.399> time,<00:16:54.720> Clint,<00:16:55.759> we<00:16:56.000> have<00:16:56.160> people<00:16:56.320> in<00:16:56.560> our<00:16:56.720> companies<00:16:57.040> that<00:16:57.279> are<00:16:57.680> building<00:16:58.240> more<00:16:58.639> software<00:17:00.320> with<00:17:00.639> AI<00:17:00.959> tools<00:17:01.440> faster<00:17:01.839> than 00:17:02.000 --> 00:17:10.400 ever,<00:17:03.040> which<00:17:03.360> unfortunately<00:17:04.319> data<00:17:04.720> shows<00:17:05.120> is<00:17:05.439> 80%<00:17:06.000> of<00:17:06.160> those<00:17:06.319> lines<00:17:06.559> of<00:17:06.720> code<00:17:06.959> being<00:17:07.120> produced<00:17:07.439> even<00:17:07.600> with<00:17:07.679> AI<00:17:08.000> coding<00:17:08.319> aliens<00:17:08.959> are<00:17:09.280> insecure. 00:17:10.400 --> 00:17:18.079 So<00:17:10.640> for<00:17:10.880> every<00:17:11.199> month<00:17:11.520> you're<00:17:11.760> letting<00:17:12.079> your<00:17:12.400> teams<00:17:12.880> use<00:17:13.039> AI<00:17:13.439> coding<00:17:13.760> assistance<00:17:14.160> to<00:17:14.400> write<00:17:14.640> code,<00:17:15.520> you're<00:17:15.760> accumulating<00:17:16.319> another<00:17:16.720> six<00:17:16.959> months<00:17:17.199> of<00:17:17.439> security<00:17:17.760> debt 00:17:18.079 --> 00:17:25.360 behind<00:17:18.314> it.<00:17:18.548> <00:17:18.783> So<00:17:19.018> you<00:17:19.253> got<00:17:19.487> to<00:17:19.722> figure<00:17:19.957> out<00:17:20.191> how<00:17:20.426> to<00:17:20.661> deal<00:17:20.895> with<00:17:21.130> the<00:17:21.365> legacy<00:17:21.599> and<00:17:21.834> you<00:17:22.069> got<00:17:22.304> to<00:17:22.538> figure<00:17:22.773> out<00:17:23.008> how<00:17:23.242> to<00:17:23.477> start<00:17:23.712> clean<00:17:23.947> with<00:17:24.181> the<00:17:24.416> new<00:17:24.651> stuff<00:17:24.885> because<00:17:25.120> this 00:17:25.360 --> 00:17:32.400 is<00:17:25.520> going<00:17:25.600> to<00:17:25.679> be<00:17:26.400> all<00:17:26.640> odds<00:17:27.039> stacked<00:17:27.360> against<00:17:27.600> you<00:17:27.760> if<00:17:28.000> you<00:17:28.079> let<00:17:28.319> people<00:17:28.640> go<00:17:28.960> build<00:17:29.280> software<00:17:29.679> and<00:17:29.840> then<00:17:29.919> you<00:17:30.080> come<00:17:30.240> back<00:17:30.400> a<00:17:30.640> year<00:17:30.799> later<00:17:31.120> saying,<00:17:31.600> "Ah,<00:17:32.000> let<00:17:32.160> me<00:17:32.320> think 00:17:32.400 --> 00:17:40.880 about<00:17:32.742> security."<00:17:33.084> Now,<00:17:33.425> <00:17:33.767> you<00:17:34.109> mentioned<00:17:34.451> data<00:17:34.792> centers<00:17:35.134> earlier<00:17:35.476> and<00:17:35.818> it<00:17:36.160> kind<00:17:36.501> of<00:17:36.843> ties<00:17:37.185> into<00:17:37.527> just<00:17:37.868> like<00:17:38.210> how<00:17:38.552> unpopular<00:17:38.894> AI<00:17:39.235> is<00:17:39.577> in<00:17:39.919> general. 00:17:40.880 --> 00:17:45.840 Like<00:17:41.039> we're<00:17:41.360> seeing<00:17:41.919> protests<00:17:42.480> and<00:17:42.720> so<00:17:42.880> many<00:17:43.039> people<00:17:43.360> protesting,<00:17:44.000> you<00:17:44.080> know,<00:17:44.240> building<00:17:44.480> of<00:17:44.720> data<00:17:44.960> centers<00:17:45.280> in<00:17:45.440> their<00:17:45.600> state. 00:17:45.840 --> 00:17:52.000 It's<00:17:46.000> happening<00:17:46.240> in<00:17:46.400> my<00:17:46.559> state.<00:17:46.799> It's<00:17:47.039> probably<00:17:47.200> happening<00:17:47.440> in<00:17:47.600> everybody's<00:17:48.000> state<00:17:48.640> where<00:17:48.960> people<00:17:49.120> are<00:17:49.280> getting<00:17:49.760> uh<00:17:49.919> freaked<00:17:50.320> out<00:17:50.640> about<00:17:50.960> these<00:17:51.280> data<00:17:51.520> centers<00:17:51.760> and 00:17:52.000 --> 00:17:54.400 stuff.<00:17:52.480> Why<00:17:52.720> do<00:17:52.799> you<00:17:52.960> think<00:17:53.120> AI<00:17:53.440> is<00:17:53.600> so<00:17:53.760> unpopular? 00:17:54.400 --> 00:17:56.880 Why<00:17:54.480> do<00:17:54.640> you<00:17:54.720> think<00:17:55.120> data<00:17:55.440> centers<00:17:55.679> are<00:17:55.840> so<00:17:56.000> unpopular? 00:17:56.880 --> 00:18:02.320 And<00:17:57.120> what<00:17:57.280> can<00:17:57.440> tech<00:17:57.760> leaders<00:17:58.160> do<00:17:58.480> to<00:17:59.840> uh<00:18:00.080> correct<00:18:00.480> that<00:18:00.799> or<00:18:01.200> to<00:18:01.600> make<00:18:01.840> that<00:18:02.080> less? 00:18:02.320 --> 00:18:10.799 So,<00:18:02.853> <00:18:03.387> look,<00:18:03.920> I'm<00:18:04.453> not<00:18:04.987> an<00:18:05.520> environmentalist,<00:18:06.053> but<00:18:06.587> I<00:18:07.120> think<00:18:07.653> logically<00:18:08.187> we<00:18:08.720> would<00:18:09.253> all<00:18:09.787> agree<00:18:10.320> that 00:18:10.799 --> 00:18:14.720 data<00:18:11.200> centers<00:18:11.440> are<00:18:11.600> going<00:18:11.679> to<00:18:11.840> draw<00:18:12.720> a<00:18:13.039> lot<00:18:13.200> of<00:18:13.440> power. 00:18:14.720 --> 00:18:18.799 They<00:18:14.960> need<00:18:15.120> a<00:18:15.520> lot<00:18:15.679> of<00:18:16.160> water<00:18:16.640> for<00:18:16.960> cooling. 00:18:18.799 --> 00:18:23.679 And<00:18:19.840> if<00:18:20.000> you<00:18:20.160> look<00:18:20.240> at<00:18:20.400> most<00:18:20.640> states,<00:18:21.200> the<00:18:21.440> grids<00:18:21.840> are,<00:18:22.400> you<00:18:22.640> know,<00:18:23.120> gridlocked. 00:18:23.679 --> 00:18:28.240 Like<00:18:23.919> we<00:18:24.080> don't<00:18:24.240> think<00:18:24.320> we<00:18:24.480> have<00:18:24.640> enough<00:18:25.039> capacity<00:18:25.520> just<00:18:25.760> to<00:18:25.919> keep<00:18:26.080> going<00:18:26.320> with,<00:18:27.440> you<00:18:27.520> know,<00:18:27.679> heat<00:18:27.919> waves. 00:18:28.240 --> 00:18:32.160 If<00:18:28.320> there's<00:18:28.480> a<00:18:28.559> heat<00:18:28.720> wave<00:18:28.960> in<00:18:29.120> California,<00:18:29.679> you<00:18:29.919> get<00:18:30.080> all<00:18:30.240> of<00:18:30.400> these<00:18:30.799> alarm<00:18:31.120> bells<00:18:31.440> from<00:18:31.679> P. 00:18:32.160 --> 00:18:34.960 <00:18:32.367> Grids<00:18:32.573> are<00:18:32.780> maxed<00:18:32.987> out.<00:18:33.193> Most<00:18:33.400> that's<00:18:33.607> something<00:18:33.813> that's<00:18:34.020> not<00:18:34.227> talked<00:18:34.433> about<00:18:34.640> enough. 00:18:34.960 --> 00:18:38.240 Like<00:18:35.120> in<00:18:35.280> in<00:18:35.520> the<00:18:35.679> United<00:18:35.840> States,<00:18:36.160> most<00:18:36.320> grids<00:18:36.640> are<00:18:36.720> maxed<00:18:37.039> out<00:18:37.200> in<00:18:37.360> most<00:18:37.520> states. 00:18:38.240 --> 00:18:43.280 <00:18:38.613> Yeah.<00:18:38.987> So<00:18:39.360> grids<00:18:39.733> are<00:18:40.107> maxed<00:18:40.480> out.<00:18:40.853> High<00:18:41.227> water<00:18:41.600> consumption,<00:18:41.973> high<00:18:42.347> cooling<00:18:42.720> needs. 00:18:43.280 --> 00:18:47.280 There's<00:18:43.520> the<00:18:43.760> carbon<00:18:44.080> footprint<00:18:44.559> of<00:18:44.799> all<00:18:44.960> of<00:18:45.120> this<00:18:46.240> uh<00:18:46.559> coming<00:18:46.799> out. 00:18:47.280 --> 00:18:56.240 And<00:18:47.440> then<00:18:47.679> there's<00:18:48.080> obviously<00:18:49.039> depending<00:18:49.360> on<00:18:49.520> who<00:18:49.760> you<00:18:49.919> talk<00:18:50.160> to<00:18:51.520> this<00:18:51.919> idea<00:18:52.480> of<00:18:53.919> um<00:18:54.640> all<00:18:54.880> of<00:18:54.960> the<00:18:55.120> electromagnetic<00:18:56.000> stuff 00:18:56.240 --> 00:18:58.000 that<00:18:56.400> comes<00:18:56.640> with<00:18:56.799> data<00:18:57.039> centers<00:18:57.360> what<00:18:57.440> it<00:18:57.520> does<00:18:57.600> to<00:18:57.760> neighborhood. 00:18:58.000 --> 00:19:06.480 So<00:18:58.311> I<00:18:58.622> don't<00:18:58.932> know<00:18:59.243> like<00:18:59.554> look<00:18:59.865> I'm<00:19:00.175> not<00:19:00.486> a<00:19:00.797> scientist<00:19:01.108> you're<00:19:01.418> going<00:19:01.729> to<00:19:02.040> describe<00:19:02.351> most<00:19:02.662> this<00:19:02.972> but<00:19:03.283> logically<00:19:03.594> <00:19:03.905> anytime<00:19:04.215> we<00:19:04.526> shift<00:19:04.837> hard<00:19:05.148> on<00:19:05.458> any<00:19:05.769> aspect<00:19:06.080> of 00:19:06.480 --> 00:19:13.679 technology<00:19:07.120> there<00:19:07.280> are<00:19:07.440> consequences<00:19:08.080> and<00:19:08.240> especially<00:19:08.480> when<00:19:08.640> you're<00:19:08.799> doing<00:19:08.880> it<00:19:09.039> in<00:19:09.200> such<00:19:09.360> a<00:19:09.520> compressed<00:19:10.480> time<00:19:10.799> frame<00:19:11.679> we<00:19:11.919> probably<00:19:12.160> haven't<00:19:12.400> studied<00:19:12.799> those<00:19:12.960> consequences 00:19:13.679 --> 00:19:19.600 right<00:19:13.840> we're<00:19:14.000> just<00:19:14.160> kind<00:19:14.320> of<00:19:14.400> scrambling<00:19:14.799> to<00:19:14.960> get<00:19:15.120> in<00:19:15.360> front<00:19:15.440> of<00:19:15.600> this<00:19:16.240> because<00:19:16.559> hey<00:19:16.799> if<00:19:16.880> if<00:19:17.200> US<00:19:17.440> doesn't<00:19:17.679> produce<00:19:17.919> those<00:19:18.160> data<00:19:18.320> centers<00:19:18.640> fast<00:19:18.799> enough<00:19:19.120> China<00:19:19.360> is<00:19:19.520> going 00:19:19.600 --> 00:19:25.440 to<00:19:19.679> win<00:19:19.840> the<00:19:20.000> race<00:19:20.160> because<00:19:20.400> they're<00:19:20.640> able<00:19:20.799> to<00:19:20.960> create<00:19:21.200> a<00:19:21.600> data<00:19:21.919> center<00:19:22.080> in<00:19:22.240> a<00:19:22.400> weekend<00:19:22.640> and<00:19:22.799> it<00:19:22.960> takes<00:19:23.120> us<00:19:23.280> two<00:19:23.520> years<00:19:23.600> to<00:19:23.760> do<00:19:23.919> it<00:19:24.000> in<00:19:24.160> the<00:19:24.320> United<00:19:24.559> States. 00:19:25.440 --> 00:19:29.360 Um,<00:19:25.840> you<00:19:26.000> know,<00:19:26.160> I<00:19:26.320> I<00:19:26.480> I<00:19:26.720> think<00:19:26.799> we're<00:19:27.039> all<00:19:27.200> kind<00:19:27.440> of<00:19:27.600> racing<00:19:28.720> down<00:19:29.039> there. 00:19:29.360 --> 00:19:37.440 I<00:19:29.520> mean,<00:19:29.679> it's<00:19:30.000> funny<00:19:30.240> to<00:19:30.400> watch<00:19:31.679> um<00:19:32.720> technology<00:19:33.440> companies<00:19:33.840> look<00:19:34.000> at<00:19:34.160> buying<00:19:34.480> or<00:19:34.720> making<00:19:34.880> investments<00:19:35.280> in<00:19:35.520> like<00:19:35.679> nuclear<00:19:36.080> power<00:19:36.320> plants<00:19:36.559> because<00:19:36.799> now<00:19:37.039> power<00:19:37.200> is<00:19:37.280> the 00:19:37.440 --> 00:19:44.400 biggest<00:19:37.679> shortage.<00:19:38.640> Uh,<00:19:39.039> look,<00:19:39.360> you<00:19:39.440> know,<00:19:39.679> they<00:19:39.840> always<00:19:40.000> say<00:19:40.160> never<00:19:40.400> bet<00:19:40.720> against<00:19:41.039> Elon<00:19:41.440> Musk,<00:19:41.760> but<00:19:41.919> like<00:19:42.240> putting<00:19:42.480> data<00:19:42.799> spanners<00:19:43.200> in<00:19:43.520> space. 00:19:44.400 --> 00:19:49.200 Incredibly<00:19:46.559> crazy<00:19:46.960> idea,<00:19:47.440> but<00:19:47.840> sounds<00:19:48.080> like<00:19:48.240> it<00:19:48.480> might<00:19:48.640> be<00:19:48.799> real. 00:19:49.200 --> 00:19:52.080 Um,<00:19:49.679> so<00:19:49.840> I<00:19:50.240> think<00:19:50.320> there's<00:19:50.559> just<00:19:50.640> a<00:19:50.799> lot<00:19:51.039> happen. 00:19:52.080 --> 00:20:00.240 It<00:19:52.320> is<00:19:52.559> funny<00:19:52.799> though,<00:19:52.960> Clint,<00:19:53.360> right?<00:19:53.760> Five<00:19:54.000> years<00:19:54.240> ago,<00:19:55.679> hardware<00:19:56.080> was<00:19:56.320> supposed<00:19:56.480> to<00:19:56.640> be<00:19:56.799> dead<00:19:57.600> and<00:19:57.840> it<00:19:58.000> was<00:19:58.080> all<00:19:58.320> about<00:19:58.640> software. 00:20:00.240 --> 00:20:08.400 Software<00:20:00.640> was<00:20:00.799> eating<00:20:01.120> the<00:20:01.280> world.<00:20:02.080> But<00:20:02.240> now<00:20:02.559> the<00:20:02.799> question<00:20:03.039> mark,<00:20:03.360> the<00:20:03.520> question<00:20:03.760> in<00:20:03.919> everybody's<00:20:04.320> mind<00:20:04.559> is<00:20:04.960> are<00:20:05.200> software<00:20:05.679> jobs<00:20:06.160> dead<00:20:07.120> because<00:20:07.440> AI<00:20:07.760> is<00:20:08.000> getting<00:20:08.160> so 00:20:08.400 --> 00:20:11.600 good<00:20:08.480> at<00:20:08.720> writing<00:20:09.039> software<00:20:09.840> and<00:20:10.080> now<00:20:10.480> hardware<00:20:10.799> is<00:20:10.960> hot<00:20:11.200> again. 00:20:11.600 --> 00:20:15.760 So<00:20:11.919> look,<00:20:12.960> we'll<00:20:13.280> have<00:20:13.440> to<00:20:13.520> see<00:20:13.679> what<00:20:13.840> the<00:20:14.080> next<00:20:14.240> year<00:20:14.320> or<00:20:14.480> two<00:20:14.720> years<00:20:14.960> play<00:20:15.120> out. 00:20:15.760 --> 00:20:23.280 There's<00:20:16.080> a<00:20:16.240> very<00:20:16.559> real<00:20:16.960> concern<00:20:17.440> right<00:20:17.679> now<00:20:18.320> as<00:20:18.559> these<00:20:18.880> frontier<00:20:19.360> labs<00:20:19.679> go<00:20:19.840> on<00:20:20.080> a<00:20:20.240> path<00:20:20.400> to<00:20:20.640> profitability,<00:20:21.280> a<00:20:21.520> path<00:20:21.679> to<00:20:22.480> going<00:20:22.799> public. 00:20:23.280 --> 00:20:31.679 The<00:20:23.440> subsidies<00:20:23.919> start<00:20:24.160> coming<00:20:24.400> off,<00:20:24.799> the<00:20:25.120> unlimited<00:20:25.679> usage<00:20:26.000> of<00:20:26.160> AI<00:20:26.960> starts<00:20:27.280> coming<00:20:27.520> off<00:20:27.840> is<00:20:28.559> well<00:20:29.120> at<00:20:29.440> what<00:20:29.840> cost<00:20:30.240> are<00:20:30.480> we<00:20:30.720> getting<00:20:31.440> this? 00:20:31.679 --> 00:20:33.600 And<00:20:31.840> I<00:20:32.000> always<00:20:32.159> say<00:20:32.240> it's<00:20:32.400> a<00:20:32.559> bit<00:20:32.640> of<00:20:32.799> like<00:20:32.960> a<00:20:33.120> drug,<00:20:33.360> right? 00:20:33.600 --> 00:20:36.080 We've<00:20:33.760> all<00:20:33.919> been<00:20:34.000> fed<00:20:34.240> this<00:20:34.400> AI<00:20:34.720> drug.<00:20:34.960> We<00:20:35.200> can't<00:20:35.360> live<00:20:35.520> without<00:20:35.840> it. 00:20:36.080 --> 00:20:43.039 Now,<00:20:36.720> I'm<00:20:36.960> sure<00:20:37.120> you<00:20:37.360> and<00:20:37.520> I<00:20:38.080> first<00:20:38.320> thing<00:20:38.480> we<00:20:38.640> do<00:20:38.880> any<00:20:39.200> question<00:20:39.440> comes<00:20:39.679> up<00:20:39.760> is<00:20:40.000> we<00:20:40.159> have<00:20:40.240> our<00:20:40.480> little<00:20:40.720> agents<00:20:41.520> assist<00:20:42.320> jobs. 00:20:43.039 --> 00:20:46.159 It<00:20:43.200> was<00:20:43.440> fine<00:20:43.760> when<00:20:44.000> we<00:20:44.240> were<00:20:44.320> just<00:20:44.480> paying<00:20:44.720> a<00:20:44.880> monthly<00:20:45.200> subscription. 00:20:46.159 --> 00:20:53.840 When<00:20:46.400> you<00:20:46.559> now<00:20:46.799> have<00:20:46.880> to<00:20:47.039> start<00:20:47.280> thinking<00:20:47.520> about<00:20:47.679> a<00:20:47.840> token,<00:20:48.640> you<00:20:48.720> know,<00:20:48.880> Uber<00:20:49.360> just<00:20:49.600> described<00:20:50.559> uh<00:20:50.720> they<00:20:50.880> have<00:20:51.039> run<00:20:51.200> out<00:20:51.280> of<00:20:51.360> their<00:20:51.520> entire<00:20:51.840> 2026<00:20:52.640> budget<00:20:53.200> in<00:20:53.440> the<00:20:53.600> first 00:20:53.840 --> 00:20:59.200 four<00:20:54.080> months<00:20:54.400> on<00:20:54.640> tokens.<00:20:55.360> It's<00:20:55.520> a<00:20:55.760> very<00:20:56.000> real<00:20:56.320> conversation<00:20:56.799> I'm<00:20:57.039> having<00:20:57.200> with<00:20:57.440> CIOs,<00:20:58.000> with<00:20:58.159> CFOs. 00:20:59.200 --> 00:21:01.360 People<00:20:59.440> don't<00:20:59.679> know<00:20:59.760> how<00:20:59.919> to<00:21:00.080> forecast<00:21:00.480> the<00:21:00.720> spend. 00:21:01.360 --> 00:21:07.360 And<00:21:01.600> most<00:21:01.760> of<00:21:01.919> this,<00:21:02.159> they're<00:21:02.640> not<00:21:02.799> able<00:21:03.039> to<00:21:03.200> kind<00:21:03.360> of<00:21:03.440> tie<00:21:03.760> to<00:21:05.360> explicit<00:21:06.000> business<00:21:06.480> value. 00:21:07.360 --> 00:21:13.440 And<00:21:07.520> so<00:21:07.679> you<00:21:07.919> build<00:21:08.080> all<00:21:08.240> of<00:21:08.400> this<00:21:08.559> capacity<00:21:09.280> because<00:21:09.600> people<00:21:09.840> are<00:21:10.080> just<00:21:10.400> using<00:21:11.440> tokens<00:21:11.919> recklessly<00:21:12.640> right<00:21:12.880> now. 00:21:13.440 --> 00:21:16.640 But<00:21:13.600> all<00:21:13.760> of<00:21:13.920> this<00:21:14.080> is<00:21:14.240> going<00:21:14.400> to<00:21:14.560> there's<00:21:14.720> going<00:21:14.880> to<00:21:15.040> be<00:21:15.200> a<00:21:15.360> day<00:21:15.520> of<00:21:15.680> reckoning<00:21:15.840> across<00:21:16.000> all<00:21:16.160> of<00:21:16.320> this. 00:21:16.640 --> 00:21:24.880 And<00:21:16.799> then<00:21:17.440> the<00:21:17.760> question<00:21:18.000> of<00:21:18.880> is<00:21:19.280> this<00:21:19.600> best<00:21:20.080> use<00:21:20.400> case<00:21:20.640> served<00:21:20.960> by<00:21:21.200> humans<00:21:22.240> or<00:21:22.480> a<00:21:22.720> small<00:21:23.039> model<00:21:23.440> or<00:21:23.679> a<00:21:23.919> large<00:21:24.159> model<00:21:24.400> like<00:21:24.640> we're 00:21:24.880 --> 00:21:32.480 going<00:21:25.039> to<00:21:25.120> have<00:21:25.280> to<00:21:25.440> get<00:21:25.600> to<00:21:25.760> that<00:21:26.559> very<00:21:26.960> quickly<00:21:27.600> because<00:21:27.840> I<00:21:28.000> think<00:21:28.159> last<00:21:28.320> 6<00:21:28.559> months<00:21:28.799> have<00:21:29.039> been<00:21:29.280> about<00:21:29.520> just<00:21:29.760> get<00:21:29.919> every<00:21:30.159> employee<00:21:30.480> to<00:21:30.640> use<00:21:30.799> AI<00:21:31.760> and<00:21:32.000> now<00:21:32.159> we're<00:21:32.400> going 00:21:32.480 --> 00:21:40.559 to<00:21:32.559> have<00:21:32.640> to<00:21:32.720> talk<00:21:32.880> about<00:21:33.120> let's<00:21:33.440> measure<00:21:33.679> what's<00:21:34.000> working<00:21:34.720> and<00:21:35.039> rightsize<00:21:35.760> AI<00:21:36.240> utilization<00:21:36.640> in<00:21:36.880> our<00:21:37.039> companies<00:21:37.679> and<00:21:37.919> that<00:21:38.080> affects<00:21:38.799> utilization<00:21:39.200> of<00:21:39.440> data<00:21:39.679> centers. 00:21:40.559 --> 00:21:49.120 <00:21:40.879> I<00:21:41.199> couldn't<00:21:41.519> agree<00:21:41.839> more.<00:21:42.159> I<00:21:42.479> think<00:21:42.799> the<00:21:43.119> uh<00:21:43.439> token<00:21:43.759> cost<00:21:44.079> is<00:21:44.399> going<00:21:44.720> to<00:21:45.040> be<00:21:45.360> the<00:21:45.680> issue<00:21:46.000> a<00:21:46.320> year<00:21:46.640> from<00:21:46.960> now<00:21:47.280> when<00:21:47.600> it's<00:21:47.920> no<00:21:48.240> longer<00:21:48.560> subsidized<00:21:48.880> maybe 00:21:49.120 --> 00:21:56.240 even<00:21:49.280> sooner<00:21:49.600> but<00:21:49.840> like<00:21:50.000> particularly<00:21:50.400> like<00:21:50.720> if<00:21:50.880> we<00:21:51.039> were<00:21:51.200> talking<00:21:51.440> again<00:21:51.840> next<00:21:52.159> June<00:21:53.520> I<00:21:53.840> think<00:21:54.400> uh<00:21:54.640> the<00:21:54.799> thing<00:21:54.960> that<00:21:55.120> everybody's<00:21:55.440> going<00:21:55.520> to<00:21:55.600> be<00:21:55.679> talking<00:21:55.919> about<00:21:56.080> I<00:21:56.159> was 00:21:56.240 --> 00:22:03.600 like<00:21:56.799> we<00:21:57.200> we<00:21:57.440> fired<00:21:57.679> a<00:21:57.919> lot<00:21:58.000> of<00:21:58.159> employees<00:21:58.559> who<00:21:59.039> were<00:21:59.280> actually<00:21:59.520> cheaper<00:21:59.919> than<00:22:00.080> the<00:22:00.480> tokens<00:22:01.600> that<00:22:01.840> could<00:22:01.919> that's<00:22:02.080> a<00:22:02.240> real<00:22:02.400> possibility<00:22:02.799> doesn't<00:22:03.039> mean<00:22:03.120> that's<00:22:03.360> definitely 00:22:03.600 --> 00:22:08.799 going<00:22:03.760> to<00:22:04.000> happen<00:22:04.240> but<00:22:04.400> that's<00:22:04.559> a<00:22:04.720> that's<00:22:04.880> a<00:22:05.039> real<00:22:05.200> possibility<00:22:05.600> you<00:22:05.679> brought<00:22:05.919> something<00:22:06.080> up<00:22:06.240> there<00:22:06.320> that<00:22:06.559> I<00:22:06.720> actually<00:22:06.880> want<00:22:06.960> to<00:22:07.120> touch<00:22:07.200> on<00:22:07.360> and<00:22:07.600> I've<00:22:07.760> been<00:22:08.159> asking<00:22:08.559> all 00:22:08.799 --> 00:22:17.280 sorts<00:22:08.960> of<00:22:09.120> CEOs<00:22:09.600> and<00:22:09.760> leaders<00:22:10.240> on<00:22:10.480> this<00:22:11.200> and<00:22:11.440> just<00:22:11.919> throughout<00:22:12.640> um<00:22:12.880> the<00:22:13.120> day<00:22:14.080> uh<00:22:14.159> what<00:22:14.400> it<00:22:14.559> means<00:22:14.799> and<00:22:14.960> you<00:22:15.200> said<00:22:15.360> something<00:22:15.520> like<00:22:15.760> if<00:22:16.080> we<00:22:16.240> don't<00:22:16.480> win<00:22:16.880> China 00:22:17.280 --> 00:22:25.760 will<00:22:17.600> win<00:22:18.000> or<00:22:18.400> some<00:22:18.640> some<00:22:18.880> something<00:22:19.280> along<00:22:19.520> the<00:22:19.679> lines<00:22:19.840> of<00:22:20.000> like<00:22:20.240> China<00:22:20.480> is<00:22:20.720> going<00:22:21.520> there's<00:22:21.840> an<00:22:22.000> AI<00:22:22.320> race<00:22:22.720> and<00:22:23.039> you're<00:22:23.360> not<00:22:23.520> saying<00:22:23.679> this<00:22:24.240> or<00:22:24.400> you<00:22:24.559> didn't<00:22:24.720> say<00:22:24.880> this<00:22:25.039> here 00:22:25.760 --> 00:22:32.320 um<00:22:26.320> what<00:22:26.559> our<00:22:26.799> leaders<00:22:27.039> are<00:22:27.200> saying<00:22:27.280> and<00:22:27.520> those<00:22:27.679> who<00:22:27.840> are<00:22:27.919> like<00:22:28.080> trying<00:22:28.320> to<00:22:28.400> build<00:22:28.559> these<00:22:28.799> data<00:22:29.039> centers<00:22:29.200> and<00:22:29.360> all<00:22:29.600> that<00:22:29.760> type<00:22:29.840> of<00:22:30.000> stuff<00:22:30.159> are<00:22:30.320> saying<00:22:31.120> is<00:22:31.919> um<00:22:32.080> there's 00:22:32.320 --> 00:22:41.360 an<00:22:32.400> AI<00:22:32.720> race<00:22:32.880> and<00:22:33.120> if<00:22:33.280> China<00:22:33.679> wins<00:22:34.159> that's<00:22:34.480> bad<00:22:35.120> and<00:22:35.520> we<00:22:35.760> need<00:22:35.919> to<00:22:36.080> win<00:22:37.120> because<00:22:37.520> why<00:22:38.240> and<00:22:38.480> what<00:22:38.799> nobody's<00:22:39.200> really<00:22:39.360> explained<00:22:39.600> is<00:22:39.760> like<00:22:39.919> what<00:22:40.240> winning<00:22:40.960> means 00:22:41.360 --> 00:22:45.919 against<00:22:41.760> China<00:22:42.640> or<00:22:42.960> really<00:22:43.200> explaining<00:22:43.760> like<00:22:43.919> that<00:22:44.159> China's<00:22:44.559> not<00:22:44.640> that<00:22:44.960> even<00:22:45.280> far<00:22:45.520> behind. 00:22:45.919 --> 00:22:50.559 So,<00:22:46.159> do<00:22:46.320> you<00:22:46.400> know<00:22:46.480> what<00:22:46.640> it<00:22:46.799> means<00:22:46.960> to<00:22:47.120> like<00:22:47.280> beat<00:22:47.520> China<00:22:47.840> in<00:22:48.000> the<00:22:48.080> AI<00:22:48.480> race? 00:22:50.559 --> 00:22:56.080 I<00:22:50.992> don't<00:22:51.426> have<00:22:51.859> a<00:22:52.293> um<00:22:52.726> I<00:22:53.160> don't<00:22:53.593> have<00:22:54.026> a<00:22:54.460> dog<00:22:54.893> in<00:22:55.327> that<00:22:55.760> race. 00:22:56.080 --> 00:23:04.880 I<00:22:56.320> would<00:22:56.480> in<00:22:56.720> particular<00:22:57.200> say<00:22:59.679> I<00:22:59.919> think<00:23:00.000> over<00:23:00.240> the<00:23:00.480> years<00:23:01.360> there's<00:23:01.679> been<00:23:01.840> this<00:23:02.159> concern<00:23:02.559> of<00:23:02.960> obviously<00:23:03.280> as<00:23:03.520> there's<00:23:03.760> been<00:23:03.919> like<00:23:04.720> you 00:23:04.880 --> 00:23:14.559 know<00:23:05.600> um<00:23:06.480> like<00:23:06.720> we<00:23:06.960> need<00:23:07.039> to<00:23:07.280> control<00:23:07.520> a<00:23:07.760> lot<00:23:07.840> of<00:23:07.919> our<00:23:08.320> supply<00:23:08.720> chain<00:23:09.120> of<00:23:10.000> things<00:23:10.240> if<00:23:10.400> you<00:23:10.480> think<00:23:10.720> like<00:23:11.039> all<00:23:11.200> of<00:23:11.440> AI<00:23:11.760> is<00:23:12.000> banking<00:23:12.400> on<00:23:13.840> hardware 00:23:14.559 --> 00:23:21.360 hardware<00:23:14.880> is<00:23:15.039> banking<00:23:15.280> on<00:23:15.520> basically<00:23:15.840> GPUs<00:23:16.559> the<00:23:16.799> wafers<00:23:17.520> for<00:23:17.760> the<00:23:17.919> most<00:23:18.080> part<00:23:18.320> come<00:23:18.480> from<00:23:18.640> Taiwan<00:23:19.440> there's<00:23:19.600> an<00:23:19.760> element<00:23:20.000> of<00:23:20.159> like<00:23:20.400> let's<00:23:20.559> make<00:23:20.720> sure<00:23:20.799> we<00:23:21.039> can<00:23:21.120> control 00:23:21.360 --> 00:23:29.200 a<00:23:21.600> lot<00:23:21.679> of<00:23:21.760> our<00:23:22.000> supply<00:23:22.240> chain<00:23:22.559> which<00:23:22.720> is<00:23:22.799> why<00:23:22.960> obviously<00:23:23.360> the<00:23:23.520> US<00:23:23.760> government<00:23:24.159> invested<00:23:24.640> in<00:23:24.799> Intel<00:23:25.200> and<00:23:25.360> is<00:23:25.520> trying<00:23:26.480> bring<00:23:26.799> more<00:23:26.960> of<00:23:27.039> that<00:23:27.280> manufacturing<00:23:27.919> back<00:23:28.640> uh<00:23:28.880> state 00:23:29.200 --> 00:23:37.280 side.<00:23:30.240> Um<00:23:30.960> I<00:23:31.200> think<00:23:31.360> my<00:23:31.679> primary<00:23:32.080> use<00:23:32.240> of<00:23:32.400> the<00:23:32.559> China<00:23:32.880> analogy<00:23:33.360> was<00:23:34.799> there<00:23:35.039> is<00:23:35.200> a<00:23:35.440> race<00:23:35.760> to<00:23:36.080> say<00:23:36.960> you<00:23:37.200> know 00:23:37.280 --> 00:23:44.000 you<00:23:37.520> do<00:23:37.679> need<00:23:38.000> compute<00:23:38.640> to<00:23:38.960> run<00:23:39.200> automation<00:23:39.679> at<00:23:40.000> scale<00:23:40.240> and<00:23:40.559> run<00:23:41.520> agents<00:23:42.080> at<00:23:42.400> scale. 00:23:44.000 --> 00:23:52.640 um<00:23:44.720> you're<00:23:45.039> going<00:23:45.200> to<00:23:45.360> need<00:23:45.520> that<00:23:45.760> capacity<00:23:46.400> created<00:23:47.120> somewhere<00:23:47.679> and<00:23:47.840> most<00:23:48.080> of<00:23:48.159> these<00:23:48.400> technology<00:23:49.120> workers<00:23:49.600> like<00:23:50.799> us<00:23:51.120> and<00:23:51.360> Silicon<00:23:51.840> Valley<00:23:52.320> is 00:23:52.640 --> 00:23:56.000 what<00:23:52.880> it<00:23:53.120> is<00:23:53.679> because<00:23:53.840> it<00:23:54.080> has<00:23:54.240> always<00:23:54.559> led<00:23:54.880> the<00:23:55.120> way. 00:23:56.000 --> 00:24:02.880 Um<00:23:56.720> in<00:23:57.120> the<00:23:57.919> AI<00:23:58.320> race,<00:23:58.640> there's<00:23:58.880> just<00:23:59.039> a<00:23:59.200> lot<00:23:59.360> of<00:23:59.520> supply<00:23:59.840> chain<00:24:00.320> dependencies<00:24:01.120> outside<00:24:01.520> of<00:24:01.679> the<00:24:01.840> United<00:24:02.159> States. 00:24:02.880 --> 00:24:08.720 But<00:24:03.039> the<00:24:03.200> most<00:24:03.440> the<00:24:03.760> most<00:24:03.919> important<00:24:04.159> thing<00:24:04.240> that<00:24:04.480> we<00:24:04.640> just<00:24:04.799> discussed<00:24:05.120> is<00:24:05.360> power<00:24:05.679> constraint,<00:24:06.159> capacity<00:24:06.640> constraint,<00:24:07.039> data<00:24:07.280> center<00:24:07.520> constraint. 00:24:08.720 --> 00:24:15.919 Um,<00:24:09.520> my<00:24:09.760> old<00:24:10.000> example<00:24:10.240> was,<00:24:10.640> I<00:24:10.799> think,<00:24:10.880> if<00:24:10.960> you<00:24:11.039> remember<00:24:11.200> during<00:24:11.520> COVID<00:24:11.840> days<00:24:12.400> in<00:24:12.640> China,<00:24:12.960> you<00:24:13.120> heard<00:24:13.360> like<00:24:13.440> they<00:24:13.679> could<00:24:13.840> put<00:24:14.000> up<00:24:14.159> a<00:24:14.320> hospital,<00:24:14.799> an<00:24:15.039> entire<00:24:15.360> infrastructure 00:24:15.919 --> 00:24:24.880 in<00:24:16.159> hospital<00:24:16.400> over<00:24:16.640> a<00:24:16.799> weekend.<00:24:18.080> And,<00:24:18.960> um,<00:24:19.440> if<00:24:19.679> you<00:24:19.840> look,<00:24:20.559> I<00:24:20.799> was<00:24:20.880> reading<00:24:21.120> an<00:24:21.360> article<00:24:21.600> last<00:24:21.919> week<00:24:22.799> where<00:24:23.120> the<00:24:24.159> automobile 00:24:24.880 --> 00:24:33.279 industry,<00:24:25.840> like<00:24:26.159> China<00:24:26.559> has<00:24:27.360> just<00:24:27.840> completely<00:24:28.480> bamboozled<00:24:29.200> everyone<00:24:29.600> else<00:24:29.919> on<00:24:30.400> EV<00:24:30.880> manufacturing,<00:24:32.799> and<00:24:33.120> you 00:24:33.279 --> 00:24:39.840 know,<00:24:33.516> you<00:24:33.753> can't<00:24:33.990> compete<00:24:34.226> on<00:24:34.463> price<00:24:34.700> when<00:24:34.937> you<00:24:35.174> get<00:24:35.411> to<00:24:35.647> manufacturing<00:24:35.884> there.<00:24:36.121> And<00:24:36.358> like,<00:24:36.595> I<00:24:36.832> think<00:24:37.068> there's<00:24:37.305> going<00:24:37.542> to<00:24:37.779> be<00:24:38.016> an<00:24:38.253> element,<00:24:38.489> the<00:24:38.726> data<00:24:38.963> center<00:24:39.200> capacity, 00:24:39.840 --> 00:24:46.880 if<00:24:40.000> you<00:24:40.159> can't<00:24:40.400> create<00:24:40.559> it<00:24:40.720> in<00:24:40.960> US,<00:24:41.919> you<00:24:42.000> know,<00:24:42.080> these<00:24:42.400> businesses<00:24:42.799> still<00:24:42.960> have<00:24:43.120> to<00:24:43.279> run,<00:24:43.520> they're<00:24:43.679> going<00:24:43.760> to<00:24:43.840> go<00:24:44.000> find<00:24:44.400> capacity<00:24:44.960> elsewhere,<00:24:45.760> and<00:24:46.000> China's<00:24:46.480> poised<00:24:46.720> to<00:24:46.799> do 00:24:46.880 --> 00:24:55.039 that<00:24:47.039> well.<00:24:47.200> So,<00:24:47.360> I<00:24:47.520> don't<00:24:47.600> think<00:24:47.760> we<00:24:47.919> can<00:24:48.000> fight<00:24:48.240> the,<00:24:48.799> let's<00:24:49.279> not<00:24:49.760> build<00:24:50.159> data<00:24:50.480> center,<00:24:50.720> but<00:24:50.880> it<00:24:51.039> was<00:24:51.200> just<00:24:52.000> as<00:24:52.240> a<00:24:52.400> competitive<00:24:52.880> advantage,<00:24:53.919> it<00:24:54.159> is<00:24:54.400> best<00:24:54.640> we<00:24:54.880> keep 00:24:55.039 --> 00:25:03.120 that<00:24:55.360> data<00:24:56.640> in<00:24:57.039> United<00:24:57.360> States<00:24:57.679> because,<00:24:57.840> given<00:24:58.159> all<00:24:58.320> of<00:24:58.400> the<00:24:58.559> other<00:24:58.640> geopolitical<00:24:59.440> concerns<00:25:00.320> and<00:25:00.559> the<00:25:00.799> concerns<00:25:01.200> generally<00:25:01.600> about<00:25:01.840> data<00:25:02.080> privacy<00:25:02.480> and<00:25:02.720> other<00:25:02.960> things 00:25:03.120 --> 00:25:10.640 in<00:25:03.360> China,<00:25:03.919> you<00:25:04.159> probably<00:25:04.400> don't<00:25:04.559> want<00:25:04.640> to<00:25:04.799> be<00:25:04.880> sitting<00:25:05.120> and<00:25:05.360> training<00:25:05.600> your<00:25:05.840> models<00:25:06.400> in<00:25:06.640> data<00:25:06.880> centers<00:25:07.200> in.<00:25:07.760> That,<00:25:07.919> that<00:25:08.320> was<00:25:08.799> the<00:25:08.960> meta<00:25:09.279> point. 00:25:10.640 --> 00:25:15.200 <00:25:10.818> Give<00:25:10.996> me<00:25:11.173> a<00:25:11.351> sense<00:25:11.529> for<00:25:11.707> what<00:25:11.884> a<00:25:12.062> typical<00:25:12.240> day<00:25:12.418> looks<00:25:12.596> like<00:25:12.773> for<00:25:12.951> you<00:25:13.129> as<00:25:13.307> CEO<00:25:13.484> of<00:25:13.662> this<00:25:13.840> company. 00:25:15.200 --> 00:25:21.279 <00:25:15.543> Yeah.<00:25:15.887> Uh,<00:25:16.230> when<00:25:16.574> I'm<00:25:16.917> not<00:25:17.261> running<00:25:17.604> around<00:25:17.948> my<00:25:18.291> six-year-old<00:25:18.635> twin<00:25:18.978> boys,<00:25:19.322> it's,<00:25:19.665> it's<00:25:20.009> usually,<00:25:20.352> it's<00:25:20.696> interesting,<00:25:21.039> right? 00:25:21.279 --> 00:25:30.159 Like,<00:25:21.600> uh,<00:25:22.559> uh,<00:25:23.120> a<00:25:23.360> few<00:25:23.520> years<00:25:23.679> ago,<00:25:23.919> I<00:25:24.080> would<00:25:24.159> say<00:25:24.240> at<00:25:24.320> the<00:25:24.480> size<00:25:24.720> and<00:25:24.960> stage<00:25:25.120> that<00:25:25.279> Endor<00:25:25.840> is,<00:25:26.880> it's<00:25:27.200> probably<00:25:28.000> a<00:25:28.240> third<00:25:28.480> on<00:25:28.880> product,<00:25:29.919> a 00:25:30.159 --> 00:25:38.480 third<00:25:30.630> on<00:25:31.100> like<00:25:31.571> go-to-market<00:25:32.041> customers,<00:25:32.512> and<00:25:32.983> a<00:25:33.453> third<00:25:33.924> on<00:25:34.394> just<00:25:34.865> our<00:25:35.335> people<00:25:35.806> and<00:25:36.277> hiring<00:25:36.747> and<00:25:37.218> continuous<00:25:37.688> scale.<00:25:38.159> I 00:25:38.480 --> 00:25:46.640 would<00:25:38.640> say<00:25:38.880> a<00:25:39.279> that<00:25:39.600> has<00:25:39.760> shifted<00:25:40.159> to<00:25:40.640> at<00:25:40.720> least<00:25:41.279> 50%<00:25:42.080> on<00:25:42.480> product,<00:25:44.159> um,<00:25:45.200> I<00:25:45.440> would<00:25:45.520> say,<00:25:45.679> because<00:25:45.840> we<00:25:46.080> live<00:25:46.159> in<00:25:46.240> a<00:25:46.480> world 00:25:46.640 --> 00:25:49.600 now<00:25:47.067> where<00:25:47.493> product-market<00:25:47.920> fit<00:25:48.347> evolves<00:25:48.773> every<00:25:49.200> quarter. 00:25:49.600 --> 00:25:52.000 It's<00:25:49.840> not<00:25:50.000> like,<00:25:50.640> it's<00:25:50.960> not<00:25:51.120> a<00:25:51.279> constant. 00:25:52.000 --> 00:25:58.640 Just<00:25:52.312> because<00:25:52.623> I<00:25:52.935> have<00:25:53.246> product-market<00:25:53.558> fit<00:25:53.869> today<00:25:54.181> doesn't<00:25:54.492> guarantee<00:25:54.804> I<00:25:55.115> have<00:25:55.427> product-market<00:25:55.738> fit<00:25:56.050> in<00:25:56.361> 6<00:25:56.673> months,<00:25:56.984> even<00:25:57.296> though<00:25:57.607> I'm<00:25:57.919> scaling. 00:25:58.640 --> 00:26:06.159 So<00:25:58.883> I<00:25:59.125> think<00:25:59.368> one<00:25:59.611> of<00:25:59.854> the<00:26:00.096> biggest<00:26:00.339> shifts<00:26:00.582> is<00:26:00.825> I<00:26:01.067> am<00:26:01.310> much<00:26:01.553> more<00:26:01.795> hands-on,<00:26:02.038> again,<00:26:02.281> reminds<00:26:02.524> me<00:26:02.766> of<00:26:03.009> the<00:26:03.252> early<00:26:03.494> days,<00:26:03.737> which<00:26:03.980> I<00:26:04.223> love,<00:26:04.465> of<00:26:04.708> Pandora,<00:26:04.951> of<00:26:05.194> like<00:26:05.436> being<00:26:05.679> deeply 00:26:06.159 --> 00:26:11.520 entrenched<00:26:06.640> in<00:26:06.799> our<00:26:07.039> product,<00:26:07.360> our<00:26:07.600> product<00:26:07.919> roadmap,<00:26:09.039> and,<00:26:09.279> and,<00:26:09.440> and<00:26:09.679> things<00:26:09.840> of<00:26:10.000> that<00:26:10.240> nature. 00:26:11.520 --> 00:26:19.840 Um,<00:26:12.720> I<00:26:12.960> think<00:26:13.039> on<00:26:13.279> hiring,<00:26:13.840> it<00:26:14.000> is<00:26:14.159> much<00:26:14.480> more<00:26:14.640> about<00:26:14.880> being<00:26:15.120> really<00:26:15.440> selective<00:26:15.919> of<00:26:16.159> getting<00:26:16.480> the<00:26:16.720> right<00:26:17.120> leaders<00:26:17.679> in<00:26:17.919> place,<00:26:18.159> because<00:26:18.480> like,<00:26:19.120> you<00:26:19.279> know,<00:26:19.360> I<00:26:19.520> argue 00:26:19.840 --> 00:26:26.320 you<00:26:20.000> don't<00:26:20.080> need<00:26:20.159> the<00:26:20.400> same<00:26:20.559> volume<00:26:20.880> of<00:26:21.039> people<00:26:21.200> in<00:26:21.440> the<00:26:21.600> company,<00:26:22.159> but<00:26:22.400> you<00:26:22.480> need<00:26:22.640> to<00:26:22.720> make<00:26:22.799> sure<00:26:22.880> you<00:26:23.039> have<00:26:23.200> the<00:26:23.360> right<00:26:23.760> mindset<00:26:24.240> of<00:26:24.480> the<00:26:24.640> leaders<00:26:25.600> that<00:26:25.840> are<00:26:26.000> finding 00:26:26.320 --> 00:26:34.640 the<00:26:26.480> right<00:26:26.640> way<00:26:26.799> to<00:26:26.960> build<00:26:27.120> efficiency,<00:26:27.840> automation,<00:26:29.360> and<00:26:29.600> just<00:26:29.840> be<00:26:30.400> way<00:26:30.799> more<00:26:31.120> willing<00:26:31.440> to<00:26:31.600> take<00:26:31.919> bets<00:26:32.240> and<00:26:32.480> risk<00:26:32.880> because<00:26:33.120> we<00:26:33.279> don't<00:26:33.440> know,<00:26:33.600> like<00:26:33.760> what<00:26:34.480> I 00:26:34.640 --> 00:26:41.760 I<00:26:35.039> commonly<00:26:35.360> have<00:26:35.520> this<00:26:35.679> problem<00:26:35.760> with<00:26:36.000> my<00:26:36.240> leaders<00:26:36.559> is,<00:26:36.720> like,<00:26:36.880> oh,<00:26:37.039> we<00:26:37.200> tried<00:26:37.440> something<00:26:37.760> six<00:26:38.000> months<00:26:38.240> ago,<00:26:39.520> it<00:26:39.760> didn't<00:26:40.000> work,<00:26:40.240> so<00:26:40.480> I'm<00:26:40.559> going<00:26:40.720> assume<00:26:41.039> when<00:26:41.200> I<00:26:41.360> sit<00:26:41.520> here 00:26:41.760 --> 00:26:50.240 today<00:26:42.240> that<00:26:42.559> it's<00:26:42.799> still<00:26:42.960> not<00:26:43.200> working,<00:26:43.440> and<00:26:43.600> I<00:26:43.760> kind<00:26:43.919> of<00:26:44.080> changing<00:26:44.400> this<00:26:45.039> mindset<00:26:45.440> of<00:26:45.679> just<00:26:45.919> because<00:26:46.480> a<00:26:46.799> use<00:26:47.039> case<00:26:47.360> of<00:26:47.760> doing<00:26:48.159> outbound<00:26:48.720> calling,<00:26:49.600> cold<00:26:49.840> calling 00:26:50.240 --> 00:26:56.960 with<00:26:50.400> AI<00:26:50.720> agents<00:26:51.120> didn't<00:26:51.440> work<00:26:51.679> 6<00:26:51.919> months<00:26:52.159> ago,<00:26:52.960> maybe<00:26:53.200> it<00:26:53.360> still<00:26:53.600> doesn't<00:26:53.840> work,<00:26:54.159> but<00:26:54.400> we<00:26:54.559> need<00:26:54.640> to<00:26:54.720> give<00:26:54.880> a<00:26:55.039> fair<00:26:55.200> shot<00:26:55.440> to<00:26:55.600> understand<00:26:55.919> how<00:26:56.080> much<00:26:56.240> that<00:26:56.480> technology<00:26:56.799> has 00:26:56.960 --> 00:27:05.200 evolved<00:26:57.360> today.<00:26:57.600> So<00:26:58.400> me<00:26:58.799> finding<00:26:59.120> the<00:26:59.440> right<00:27:00.559> mindset<00:27:00.960> of<00:27:01.200> leaders<00:27:01.520> that<00:27:01.679> we<00:27:01.840> hired<00:27:02.080> into<00:27:02.240> the<00:27:02.400> company,<00:27:02.559> I<00:27:02.799> spend<00:27:02.960> a<00:27:03.120> lot<00:27:03.200> more<00:27:03.440> time<00:27:03.600> on,<00:27:04.640> uh,<00:27:04.799> but<00:27:04.960> yeah, 00:27:05.200 --> 00:27:13.200 like<00:27:05.360> I<00:27:05.520> think<00:27:05.600> the<00:27:05.760> single<00:27:06.000> biggest<00:27:06.320> shift<00:27:06.480> is<00:27:06.720> way<00:27:06.960> more<00:27:07.200> time<00:27:07.360> on<00:27:07.679> product,<00:27:08.400> product<00:27:08.799> strategy,<00:27:09.200> and,<00:27:09.440> and<00:27:09.600> some<00:27:09.679> of<00:27:09.760> the<00:27:09.919> technology<00:27:10.559> partnerships,<00:27:12.000> um,<00:27:12.559> you<00:27:12.720> know,<00:27:12.960> in 00:27:13.200 --> 00:27:21.520 our<00:27:13.360> case,<00:27:13.679> for<00:27:13.760> example,<00:27:14.320> everything<00:27:14.559> we<00:27:14.720> do<00:27:14.880> is<00:27:15.039> around<00:27:15.360> coding<00:27:16.240> and,<00:27:16.799> uh,<00:27:16.960> protecting<00:27:17.279> the<00:27:17.440> safety<00:27:17.840> harnesses<00:27:18.320> around<00:27:18.559> coding.<00:27:18.960> So<00:27:19.600> you<00:27:19.760> know,<00:27:19.840> whether<00:27:20.080> it's<00:27:20.320> Cursor,<00:27:21.039> Anthropic, 00:27:21.520 --> 00:27:30.480 OpenAI,<00:27:22.320> Google,<00:27:23.679> Microsoft,<00:27:24.320> spending<00:27:24.799> time<00:27:25.279> in<00:27:25.600> those<00:27:25.840> technology<00:27:26.320> partnerships<00:27:26.799> is,<00:27:26.960> is<00:27:27.440> hugely<00:27:27.840> valuable<00:27:28.240> because<00:27:28.480> that<00:27:28.720> creates<00:27:29.039> the<00:27:29.840> asymmetric, 00:27:30.480 --> 00:27:38.159 nonlinear<00:27:30.822> growth<00:27:31.164> trajectory<00:27:31.505> if<00:27:31.847> we<00:27:32.189> can<00:27:32.531> unlock<00:27:32.873> better<00:27:33.215> go-to-market<00:27:33.556> with<00:27:33.898> our<00:27:34.240> partners.<00:27:34.582> You<00:27:34.924> mentioned<00:27:35.265> the<00:27:35.607> importance<00:27:35.949> of<00:27:36.291> leadership,<00:27:36.633> and<00:27:36.975> as,<00:27:37.316> as<00:27:37.658> you<00:27:38.000> were 00:27:38.159 --> 00:27:48.000 saying<00:27:38.240> that,<00:27:38.480> I<00:27:38.720> wonder<00:27:38.960> what<00:27:39.120> you<00:27:39.360> think<00:27:39.520> of<00:27:39.919> like<00:27:40.080> some<00:27:40.320> of<00:27:40.400> the<00:27:40.640> core<00:27:40.960> principles<00:27:41.279> of<00:27:41.520> leadership<00:27:42.000> are,<00:27:42.720> and<00:27:43.039> maybe<00:27:43.200> even<00:27:43.440> like<00:27:43.679> how<00:27:43.840> you<00:27:44.000> empower<00:27:44.640> first-time<00:27:45.120> leaders. 00:27:48.000 --> 00:27:56.240 <00:27:48.353> Yeah.<00:27:48.705> I<00:27:49.058> look,<00:27:49.411> I've<00:27:49.764> been<00:27:50.116> a<00:27:50.469> big<00:27:50.822> believer<00:27:51.175> in<00:27:51.527> giving<00:27:51.880> people<00:27:52.233> that<00:27:52.585> have<00:27:52.938> demonstrated<00:27:53.291> the<00:27:53.644> right<00:27:53.996> skills<00:27:54.349> the<00:27:54.702> chance<00:27:55.055> to<00:27:55.407> keep<00:27:55.760> scaling. 00:27:56.240 --> 00:28:04.880 Like,<00:27:56.686> you<00:27:57.133> know,<00:27:57.579> commonly<00:27:58.025> in<00:27:58.472> startups,<00:27:58.918> actually,<00:27:59.364> there's<00:27:59.811> data<00:28:00.257> that<00:28:00.703> shows<00:28:01.149> most<00:28:01.596> leaders<00:28:02.042> at<00:28:02.488> startups,<00:28:02.935> growth<00:28:03.381> startups,<00:28:03.827> high-growth<00:28:04.274> startups<00:28:04.720> have 00:28:04.880 --> 00:28:13.279 an<00:28:05.120> average<00:28:05.360> tenure<00:28:05.679> of<00:28:05.919> 18<00:28:06.240> months,<00:28:07.440> and<00:28:08.320> my<00:28:08.640> default<00:28:09.039> position<00:28:09.440> is<00:28:09.840> I'm<00:28:10.000> going<00:28:10.080> to<00:28:10.240> give<00:28:10.480> leaders<00:28:10.880> the<00:28:11.120> chance<00:28:11.279> to<00:28:11.520> keep<00:28:11.760> scaling<00:28:12.159> until,<00:28:12.640> like,<00:28:13.039> I 00:28:13.279 --> 00:28:18.399 have<00:28:13.520> reason<00:28:14.559> to,<00:28:14.799> to<00:28:15.039> see<00:28:15.200> or<00:28:15.440> evidence<00:28:15.840> to<00:28:16.080> believe<00:28:16.320> that<00:28:17.120> they<00:28:17.440> can<00:28:17.600> no<00:28:17.760> longer<00:28:18.080> scale. 00:28:18.399 --> 00:28:22.880 So<00:28:18.579> I'm,<00:28:18.759> I'm,<00:28:18.939> I'm<00:28:19.119> never<00:28:19.299> the,<00:28:19.479> "Hey,<00:28:19.659> you've<00:28:19.839> never<00:28:20.019> done<00:28:20.200> this<00:28:20.380> scale,<00:28:20.560> so<00:28:20.740> I'm<00:28:20.920> not<00:28:21.100> going<00:28:21.280> to<00:28:21.460> give<00:28:21.640> you<00:28:21.820> a<00:28:22.000> chance." 00:28:22.880 --> 00:28:27.919 Um,<00:28:23.760> I<00:28:24.000> also<00:28:24.320> think<00:28:24.559> leaders<00:28:25.120> in<00:28:25.360> the<00:28:25.600> AI<00:28:26.000> era<00:28:26.240> today<00:28:26.480> need<00:28:26.640> to<00:28:26.799> be<00:28:26.960> more<00:28:27.120> hands-on. 00:28:27.919 --> 00:28:29.919 I<00:28:28.159> think<00:28:28.240> there's<00:28:28.480> just<00:28:28.640> general<00:28:28.960> compression<00:28:29.360> of<00:28:29.520> skill<00:28:29.679> set. 00:28:29.919 --> 00:28:38.159 So<00:28:30.080> the<00:28:30.320> people<00:28:30.399> that<00:28:30.640> were<00:28:30.799> very<00:28:31.120> like,<00:28:31.600> "Oh,<00:28:31.919> let<00:28:32.080> me<00:28:32.240> as<00:28:32.399> I<00:28:32.559> scale<00:28:32.799> get<00:28:33.039> more<00:28:33.279> strategic<00:28:33.760> visionaries,"<00:28:35.200> they're<00:28:35.520> great,<00:28:35.679> but<00:28:35.840> I<00:28:36.000> need<00:28:36.159> visionaries<00:28:36.640> who<00:28:36.799> can<00:28:36.960> get<00:28:37.279> done. 00:28:38.159 --> 00:28:47.600 So,<00:28:38.658> um,<00:28:39.157> you<00:28:39.656> know,<00:28:40.155> you're<00:28:40.653> compressing<00:28:41.152> between<00:28:41.651> execution<00:28:42.150> and<00:28:42.649> strategy,<00:28:43.148> um,<00:28:43.647> to<00:28:44.146> kind<00:28:44.644> of<00:28:45.143> get<00:28:45.642> more,<00:28:46.141> uh,<00:28:46.640> multi-talented 00:28:47.600 --> 00:28:51.200 people,<00:28:48.000> bless<00:28:48.240> you,<00:28:48.799> that<00:28:49.039> can<00:28:49.200> kind<00:28:49.360> of<00:28:49.520> do,<00:28:49.840> do<00:28:50.159> both. 00:28:51.200 --> 00:28:58.799 Um,<00:28:52.080> the<00:28:52.399> third<00:28:52.640> thing<00:28:52.799> is<00:28:53.360> you<00:28:53.600> need<00:28:53.760> extremely<00:28:54.240> adaptable<00:28:54.880> leaders<00:28:55.200> because<00:28:55.440> a<00:28:55.679> lot<00:28:55.760> of<00:28:55.919> people<00:28:56.080> that,<00:28:56.240> that<00:28:56.640> are<00:28:56.799> the<00:28:56.960> playbook<00:28:57.440> people,<00:28:57.679> like,<00:28:57.919> "Hey,<00:28:58.080> I've<00:28:58.320> done<00:28:58.480> this 00:28:58.799 --> 00:29:07.039 before,<00:29:00.399> and<00:29:00.640> because<00:29:00.880> it<00:29:01.120> worked<00:29:01.360> for<00:29:01.520> the<00:29:01.600> last<00:29:01.840> 15<00:29:02.240> years,<00:29:02.720> I'm<00:29:02.960> going<00:29:03.039> to<00:29:03.200> keep<00:29:03.440> doing<00:29:03.760> that,"<00:29:04.960> um,<00:29:05.360> will<00:29:05.600> fail<00:29:05.919> in<00:29:06.159> this<00:29:06.320> world<00:29:06.480> we<00:29:06.720> live<00:29:06.799> in. 00:29:07.039 --> 00:29:15.200 You<00:29:07.200> have<00:29:07.360> to<00:29:07.520> be<00:29:07.679> willing<00:29:07.919> to<00:29:08.159> unlearn<00:29:09.279> a<00:29:09.520> lot<00:29:09.600> of<00:29:09.760> things<00:29:09.919> that<00:29:10.159> you<00:29:10.399> have<00:29:11.440> learned<00:29:11.760> and<00:29:12.000> developed<00:29:12.399> expertise<00:29:12.960> around,<00:29:13.440> cuz<00:29:13.679> a<00:29:13.919> lot<00:29:14.080> of<00:29:14.159> your<00:29:14.320> assumptions<00:29:14.799> about<00:29:15.039> how 00:29:15.200 --> 00:29:21.760 the<00:29:15.360> world<00:29:15.600> works,<00:29:16.000> what<00:29:16.240> technology<00:29:16.799> stacks<00:29:17.279> work,<00:29:18.640> what<00:29:18.880> kind<00:29:18.960> of<00:29:19.120> people<00:29:19.279> you<00:29:19.440> want<00:29:19.600> to<00:29:19.679> hire<00:29:20.080> is<00:29:20.320> shifting. 00:29:21.760 --> 00:29:30.000 Um,<00:29:22.799> you<00:29:23.039> know,<00:29:23.919> just<00:29:24.159> one<00:29:24.399> example<00:29:24.720> of<00:29:24.960> the,<00:29:25.120> the,<00:29:25.200> the<00:29:25.600> kind<00:29:25.840> of<00:29:27.200> mental<00:29:27.679> stimulation<00:29:28.240> exercises<00:29:28.799> I,<00:29:28.960> I<00:29:29.279> work<00:29:29.440> with<00:29:29.600> my<00:29:29.760> leaders 00:29:30.000 --> 00:29:36.960 on<00:29:30.880> is<00:29:31.679> right<00:29:31.840> now,<00:29:32.080> even<00:29:32.240> though<00:29:32.399> we're<00:29:32.559> scaling<00:29:32.960> rapidly<00:29:33.279> as<00:29:33.440> a<00:29:33.679> company,<00:29:34.240> it's<00:29:34.480> a<00:29:34.720> default<00:29:35.120> you're<00:29:35.360> not<00:29:35.520> getting<00:29:35.679> headcount. 00:29:36.960 --> 00:29:45.360 Not<00:29:37.200> to<00:29:37.360> say<00:29:37.520> we'll<00:29:37.760> never<00:29:37.919> hire,<00:29:38.320> but<00:29:38.559> it's<00:29:38.799> like<00:29:38.960> I<00:29:39.120> want<00:29:39.200> you<00:29:39.360> to<00:29:39.520> critically<00:29:39.919> think<00:29:40.159> and<00:29:40.399> tell<00:29:40.559> me<00:29:41.679> how<00:29:42.000> your<00:29:42.159> job<00:29:42.399> roles<00:29:42.720> in<00:29:42.880> your<00:29:43.039> team<00:29:43.279> are<00:29:43.520> changing,<00:29:44.559> right? 00:29:45.360 --> 00:29:48.880 Where<00:29:45.679> is<00:29:45.840> it<00:29:46.080> that<00:29:46.240> you<00:29:46.399> get<00:29:46.559> more<00:29:46.799> efficiency<00:29:47.279> out<00:29:47.440> of<00:29:47.520> the<00:29:47.760> same<00:29:47.919> people? 00:29:48.880 --> 00:29:53.279 Where<00:29:49.120> is<00:29:49.279> it<00:29:49.440> that<00:29:49.679> we<00:29:49.840> need<00:29:50.000> to<00:29:50.399> rewrite<00:29:51.679> job<00:29:52.000> descriptions? 00:29:53.279 --> 00:29:57.520 Um,<00:29:53.919> if<00:29:54.159> we<00:29:54.399> just<00:29:54.559> in<00:29:54.720> a<00:29:54.880> high<00:29:55.039> growth<00:29:55.279> mind<00:29:55.600> keep<00:29:55.760> going<00:29:55.919> and<00:29:56.159> saying,<00:29:56.240> "Oh,<00:29:56.480> I've<00:29:56.640> been<00:29:56.720> hiring<00:29:56.960> these<00:29:57.200> people. 00:29:57.520 --> 00:30:04.720 Let<00:29:57.600> me<00:29:57.760> just<00:29:57.919> keep<00:29:58.080> hiring<00:29:58.320> the<00:29:58.480> same<00:29:58.720> profile.<00:30:00.159> You<00:30:00.240> know,<00:30:00.480> you're<00:30:00.720> not<00:30:00.799> going<00:30:00.880> to<00:30:01.039> kind<00:30:01.200> of<00:30:01.360> change<00:30:01.600> your<00:30:02.399> trajectory<00:30:02.880> as<00:30:03.039> a<00:30:03.279> company<00:30:03.520> to<00:30:03.760> be<00:30:03.919> more<00:30:04.080> efficient." 00:30:04.720 --> 00:30:08.159 So,<00:30:05.279> I<00:30:05.440> think<00:30:05.600> you<00:30:05.760> definitely<00:30:05.919> need<00:30:06.159> people<00:30:06.320> that<00:30:06.480> are<00:30:06.559> much<00:30:06.720> more<00:30:06.880> adaptable. 00:30:08.159 --> 00:30:15.679 Um,<00:30:09.200> people<00:30:09.360> that<00:30:09.600> are<00:30:09.760> hands-on<00:30:10.240> yet<00:30:10.480> can<00:30:10.640> think<00:30:10.880> strategically,<00:30:11.600> which<00:30:11.840> means<00:30:12.320> by<00:30:12.559> nature<00:30:12.880> you're<00:30:13.120> giving<00:30:13.279> people<00:30:13.520> a<00:30:13.679> chance<00:30:13.840> to<00:30:14.000> keep<00:30:14.240> growing<00:30:14.480> with<00:30:14.720> you<00:30:15.279> versus 00:30:15.679 --> 00:30:18.880 bringing<00:30:15.919> in<00:30:16.159> highly<00:30:16.480> strategic<00:30:17.039> people<00:30:17.200> from<00:30:17.440> the<00:30:17.520> outside. 00:30:18.880 --> 00:30:26.159 You<00:30:19.200> mentioned<00:30:19.440> chance,<00:30:19.760> and<00:30:19.919> that's<00:30:20.159> actually<00:30:20.320> the<00:30:20.559> final<00:30:20.880> question<00:30:21.120> we<00:30:21.440> ask<00:30:21.679> everybody,<00:30:22.720> uh,<00:30:23.360> when,<00:30:23.520> when<00:30:23.840> we<00:30:24.080> interview<00:30:24.320> them,<00:30:24.559> and<00:30:24.720> that<00:30:24.880> is<00:30:25.120> we<00:30:25.360> believe<00:30:25.600> the<00:30:25.840> chances 00:30:26.159 --> 00:30:29.360 one<00:30:26.480> gives<00:30:26.960> just<00:30:27.200> important<00:30:27.440> as<00:30:27.600> the<00:30:27.760> chances<00:30:28.080> one<00:30:28.320> takes. 00:30:29.360 --> 00:30:33.919 When<00:30:29.600> you<00:30:29.760> hear<00:30:29.919> that,<00:30:30.240> who<00:30:30.399> gave<00:30:30.559> you<00:30:30.720> a<00:30:30.880> chance<00:30:30.960> to<00:30:31.120> get<00:30:31.279> you<00:30:31.360> to<00:30:31.520> where<00:30:31.679> you<00:30:31.840> are<00:30:32.080> today? 00:30:33.919 --> 00:30:39.279 Um,<00:30:34.399> I<00:30:34.720> remember<00:30:34.960> coming<00:30:35.200> out<00:30:35.360> of<00:30:35.520> college<00:30:36.000> back<00:30:36.159> in<00:30:36.399> 2006,<00:30:37.120> coming<00:30:37.279> out<00:30:37.440> of<00:30:37.520> undergrad<00:30:38.320> in<00:30:38.559> computer<00:30:38.880> science. 00:30:39.279 --> 00:30:47.440 I,<00:30:39.600> I<00:30:39.919> studied<00:30:40.320> computer<00:30:40.640> science.<00:30:40.880> I<00:30:41.200> just<00:30:43.120> did<00:30:43.440> not<00:30:45.120> enjoy<00:30:46.159> programming<00:30:47.039> all<00:30:47.279> day 00:30:47.440 --> 00:30:49.520 and<00:30:47.600> night.<00:30:47.840> I<00:30:48.000> love<00:30:48.159> being<00:30:48.320> on<00:30:48.480> the<00:30:48.559> business<00:30:48.799> side<00:30:48.960> of<00:30:49.120> technology. 00:30:49.520 --> 00:30:53.840 I<00:30:49.888> ended<00:30:50.256> up<00:30:50.624> taking<00:30:50.992> an<00:30:51.360> elective<00:30:51.728> course<00:30:52.096> on<00:30:52.464> cybersecurity<00:30:52.832> in<00:30:53.200> undergrad. 00:30:53.840 --> 00:30:57.360 Loved<00:30:54.080> it.<00:30:54.240> I'm<00:30:54.399> like,<00:30:54.480> "Holy,<00:30:54.960> you<00:30:55.120> can<00:30:55.200> like<00:30:55.440> break<00:30:55.600> into<00:30:55.840> people's<00:30:56.159> systems<00:30:56.399> and<00:30:56.640> get<00:30:56.720> paid<00:30:56.960> for<00:30:57.120> it. 00:30:57.360 --> 00:31:06.480 That's<00:30:58.000> cool."<00:30:58.799> Uh,<00:30:59.200> the<00:30:59.440> only<00:30:59.600> problem<00:30:59.840> was<00:31:00.080> in<00:31:00.320> 2006,<00:31:00.960> nobody<00:31:01.279> really<00:31:01.520> hired<00:31:02.159> kids<00:31:02.480> out<00:31:02.640> of<00:31:02.799> college<00:31:03.600> for,<00:31:04.080> for,<00:31:04.720> uh,<00:31:05.279> information<00:31:05.840> security. 00:31:06.480 --> 00:31:09.200 You<00:31:06.720> typically<00:31:07.120> were<00:31:07.360> several<00:31:07.600> years<00:31:07.840> in<00:31:08.080> it<00:31:08.320> to<00:31:08.480> become<00:31:08.720> that. 00:31:09.200 --> 00:31:17.679 And<00:31:09.360> so<00:31:09.600> long<00:31:09.760> story<00:31:10.080> short,<00:31:11.120> um,<00:31:11.760> you<00:31:11.919> know,<00:31:12.080> I<00:31:12.399> got<00:31:12.559> a<00:31:12.799> chance<00:31:13.120> at,<00:31:13.919> um,<00:31:14.799> at<00:31:15.039> KPMG<00:31:16.640> to,<00:31:17.520> uh, 00:31:17.679 --> 00:31:20.240 you<00:31:17.760> know,<00:31:17.840> I<00:31:18.000> took<00:31:18.080> a<00:31:18.240> job<00:31:18.320> at<00:31:18.480> KPMG<00:31:19.039> in<00:31:19.200> consulting. 00:31:20.240 --> 00:31:28.399 I<00:31:20.573> really<00:31:20.907> kind<00:31:21.240> of<00:31:21.573> weaseled<00:31:21.907> my<00:31:22.240> way<00:31:22.573> in<00:31:22.907> to<00:31:23.240> say,<00:31:23.573> like,<00:31:23.907> I<00:31:24.240> really<00:31:24.573> want<00:31:24.907> to<00:31:25.240> do<00:31:25.573> information<00:31:25.907> security-oriented<00:31:26.240> projects,<00:31:26.573> and,<00:31:26.907> uh,<00:31:27.240> you<00:31:27.573> know,<00:31:27.907> there<00:31:28.240> was 00:31:28.399 --> 00:31:36.720 a<00:31:28.720> manager<00:31:29.120> named<00:31:29.360> Vijay<00:31:29.840> Jaju<00:31:31.039> who,<00:31:31.600> um,<00:31:32.000> kind<00:31:32.240> of<00:31:32.720> let<00:31:33.039> me,<00:31:33.279> a<00:31:33.440> little<00:31:33.600> kid,<00:31:33.840> along<00:31:34.159> on<00:31:34.559> projects<00:31:35.039> at<00:31:35.279> Fortune<00:31:35.840> 500s<00:31:36.480> on 00:31:36.720 --> 00:31:44.960 cybersecurity,<00:31:37.071> and,<00:31:37.423> uh,<00:31:37.774> you<00:31:38.125> know,<00:31:38.476> the<00:31:38.828> rest<00:31:39.179> is<00:31:39.530> history.<00:31:39.881> Incredible,<00:31:40.233> and<00:31:40.584> we<00:31:40.935> continue<00:31:41.286> to<00:31:41.638> stay<00:31:41.989> in<00:31:42.340> touch.<00:31:42.691> Thank<00:31:43.043> you<00:31:43.394> so<00:31:43.745> much<00:31:44.096> for<00:31:44.448> coming<00:31:44.799> on 00:31:44.960 --> 00:31:52.559 the<00:31:45.039> show,<00:31:45.279> seriously.<00:31:45.600> You're<00:31:45.760> working<00:31:45.919> on<00:31:46.080> the,<00:31:46.559> I<00:31:46.799> think<00:31:46.960> maybe<00:31:47.279> the<00:31:47.600> most<00:31:47.840> important<00:31:48.320> issue,<00:31:49.279> certainly<00:31:49.600> the<00:31:49.760> most<00:31:50.000> important<00:31:50.320> vulnerability<00:31:51.120> in,<00:31:51.360> in<00:31:51.600> the<00:31:51.760> age<00:31:52.000> of<00:31:52.159> AI. 00:31:52.559 --> 00:31:55.840 So,<00:31:52.720> thank<00:31:52.880> you<00:31:52.960> for<00:31:53.120> what<00:31:53.279> you're<00:31:53.440> doing,<00:31:53.600> and<00:31:54.399> appreciate<00:31:54.720> you<00:31:54.880> coming<00:31:55.039> on. 00:31:55.840 --> 00:31:58.799 <00:31:56.080> Thanks,<00:31:56.320> Clint.<00:31:56.559> Great<00:31:56.799> chatting. ```