AI Companies Still Haven’t Delivered on Their Biggest Promises
The episode covers ZAI's release of GLM 5.3, Anthropic's unreleased advanced models, and a significant public exchange between Anthropic CEO Dario Amadei and investor Gavin Baker over claims that Anthropic leaders believe they might be the only private company left, touching on regulatory approaches, messaging strategy, and trust in AI.
Summary
The episode begins with AI model news, focusing on ZAI's release of GLM 5.3, a Chinese open-weight model that shows competitive performance improvements through scaling reinforcement learning. GLM 5.3 scores slightly behind frontier models like GPT-5.6 and Fable 5 on coding tasks but achieves state-of-the-art results on agentic benchmarks like AutomationBench and GDPVal. Notably, it surpasses Fable 5 on cybersecurity benchmarks, prompting ZAI to argue that defensive AI capabilities shouldn't be limited to closed-source companies. The host clarifies that scoring well on cyber vulnerability detection doesn't equate to autonomous attack capabilities. Wall Street analysts are increasingly recognizing that Chinese labs are genuinely competitive rather than simply using distillation or benchmark-maxing, and that cost advantages have narrowed considerably. Cloud infrastructure remains essential even for open-weight models, supporting GPU investment.
AnthropicNews reveals the company has three unreleased internal models: Opus 5, Model 1 (comparable to Mythos 5), and Model 2 (notably more capable than Mythos 5). Model 2 scores 62.8% on Anthropic's internal benchmark versus 50.3% for Mythos 5, suggesting a significant gap between public and internal capabilities. The report is already a month old as of recording, indicating further progress likely. Anthropic reportedly won't release their next model but will keep it for internal use. OpenAI staff have begun suggesting Astra will soon be available. In IPO preparations, Anthropic disclosed 14x revenue growth reaching $11.5 billion annualized run rate, with investors expecting $2 trillion valuation and $100-120 billion revenue by end of year, though Anthropic itself projects $190-200 billion by 2028.
The main segment focuses on a public discourse event initiated when investor Gavin Baker mentioned on the All In podcast that he'd heard from multiple Anthropic sources that company leadership believes Anthropic might eventually be the only private company left. After Sholto Douglas (Anthropic co-founder) disputed this as false on social media, Dario Amadei himself responded with detailed posts—a rare public engagement from someone who typically avoids social media.
Dario's response addressed two main critiques. On regulation, he rejected the framing that regulation inherently equals regulatory capture and concentration. He argued this is overly simplified and that institutional processes can actually decentralize power by vesting it in ideas rather than people. He cited Anthropic's policy proposals as intentionally disadvantaging frontier companies while helping competitors and open-weight models. He expressed support for Trump administration approaches involving pre-deployment testing and FINRA-like entities, claiming these represent progress from the previous industry consensus against regulation.
On messaging, Dario disputed claims that his communication has been disproportionately negative, citing one essay each on risks and benefits. He acknowledged that social media clips tend to focus on negative soundbites but argued the real problem isn't his messaging but rather a fundamental trust crisis in institutions spanning decades. He rejected the notion that glossy marketing will solve AI trust issues, arguing instead that actions matter: curing disease, not just talking about it. He stated the core criticism of AI companies is failing to deliver on promises to benefit humanity, and that Anthropic is ramping up biology and medicine efforts to produce real results.
Reactions were mixed. Some praised Dario for thoughtful engagement and transparency. Others critiqued his claim of balanced messaging, noting he measures balance by counting essays rather than understanding actual media dynamics and how soundbites function in public discourse. Some questioned whether he actually addressed Gavin's original claim. Debate erupted over whether AI structurally concentrates power (Dario's position) or whether compute costs and efficiency improvements will eventually distribute it. Critics noted that even significant AI breakthroughs like pharmaceutical advances won't automatically rebuild trust if companies maintain opacity around pricing, access, and power distribution. The host concludes that while nothing was definitively resolved, the public nature of the debate improved discourse quality and demonstrated the value of Dario's medium-length social media posts over longer essays for public engagement.
About this episode
<p>Anthropic CEO Dario Amodei says the strongest criticism of AI companies is that they still haven’t delivered the enormous benefits they’ve promised—and that no amount of marketing can substitute for real results. His rare public response sparks a larger debate over what the industry must actually do to prove its value. In the headlines: ZAI releases GLM 5.3, Anthropic keeps a powerful new model internal, and investors anticipate a $2 trillion Anthropic IPO.</p><p><strong>AIDB's AI Summer Adventure:</strong> <a href="https://summeradventure.ai/">https://summeradventure.ai/</a></p><p><strong>Brought to you by:</strong></p><p><strong>KPMG</strong> – Research from KPMG and the University of Texas at Austin shows the highest-impact AI users treat AI like a reasoning partner — and those skills can be taught at scale. Learn more at <a href="https://kpmg.com/us/Sophisticated">https://kpmg.com/us/Sophisticated</a></p><p><strong>Harbor - </strong>Invest in the AI ecosystem. <a href="https://www.harborcapital.com/aidaily">https://www.harborcapital.com/aidaily</a></p><p><strong>Hyperagent </strong>-<strong> </strong>Hire a fleet of always-on agents. New users get $1,000 in inference. <a href="https://hyperagent.com/aidailybrief">hyperagent.com/aidailybrief</a></p><p><strong>Rackspace Technology-</strong> One accountable partner to build, operate and run your full enterprise AI stack <a href="https://www.rackspace.com/">https://www.rackspace.com/</a></p><p><strong>Section</strong> - Section turns AI investment into workforce transformation and ROI - <a href="https://www.sectionai.com/">https://www.sectionai.com/</a></p><p><strong>Blitzy - </strong>Want to accelerate enterprise software development velocity by 5x? <a href="https://blitzy.com/">https://blitzy.com/</a></p><p><strong>AssemblyAI</strong> - The best way to build Voice AI apps - <a href="https://www.assemblyai.com/brief">https://www.assemblyai.com/brief</a></p><p><strong>Robots & Pencils</strong> - Cloud-native AI solutions that power results <a href="https://robotsandpencils.com/">https://robotsandpencils.com/</a></p><p>The AI Daily Brief helps you understand the most important news and discussions in AI. Subscribe to the podcast version of The AI Daily Brief wherever you listen: <a href="https://pod.link/1680633614">https://pod.link/1680633614</a></p><p><strong>Our Newsletter is BACK: </strong><a href="https://aidailybrief.beehiiv.com/">https://aidailybrief.beehiiv.com/</a></p><p><strong>Interested in sponsoring the show? </strong>[email protected]</p><p><br /></p>
Key Insights
- ZAI achieved significant cybersecurity benchmark improvements on GLM 5.3 through scaling reinforcement learning rather than model size, arguing that defensive AI capabilities shouldn't be monopolized by closed-source companies.
- Anthropic has three unreleased internal models with Model 2 performing 24% better than publicly-available Mythos 5 on internal benchmarks, indicating a substantial capability gap between what labs use internally versus what they release publicly.
- Dario Amadei rejected the common Silicon Valley view that all regulation inherently leads to regulatory capture, arguing instead that well-designed institutional processes can decentralize power by vesting it in ideas rather than people.
- Dario argued that Anthropic's proposed regulations intentionally disadvantage frontier companies while helping competitors and open-weight models, contrary to suggestions that Anthropic's regulatory advocacy serves its own interests.
- Dario contended that the public trust crisis with AI companies is fundamentally rooted in decades of institutional distrust across society, not primarily caused by warnings about AI risks from industry leaders.
- Dario asserted that traditional marketing campaigns claiming AI will cure cancer now fail to inspire trust because such claims have become clichés and are viewed as deceptive without demonstrated results.
- Dario stated that the most accurate criticism of AI companies including Anthropic is their failure to deliver on promises to benefit humanity, explicitly accepting this as valid rather than deflecting.
- Multiple critics argued that even major AI breakthroughs won't rebuild public trust unless companies address concerns about pricing, access, power distribution, economic gains distribution, and who ultimately holds decision-making authority.
Topics
Transcript
In a recent podcast appearance, a prominent investor said that he had heard from multiple sources inside Anthropic that Dario Amadei and other leaders in that company felt that at some point in the future, they might be the only company left. It would just be them, governments, and the rest of us. Now, these comments on that podcast kicked off quite a firestorm of discourse about Anthropic and their role in AI and what their beliefs actually meant for the industry. It also generated that rarest of phenomenon, an appearance on social media from Anthropic CEO Dario Amadei himself. In his response post, Dario discusses his real views on regulatory capture, what he thinks the real root of AI's…
Full transcript available for MurmurCast members
Sign Up to AccessMore from The AI Daily Brief: Artificial Intelligence News and Analysis
The New Problems AI Is Creating (And How People Are Solving Them)
The AI landscape has shifted from debating whether AI will be significant to solving concrete organizational challenges like token costs, AI-generated content quality, and workforce skill development. Companies are now implementing sophisticated governance structures, AI writing policies, and balanced investment strategies rather than treating AI as a simple software purchase.
How to Decide What Work AI Should Do for You: The AI Deputization Audit
The episode explores new AI capabilities (Grokbot's teach-a-task and ChatGPT's Computer History) that allow AI to learn user workflows, then introduces the AI Deputization Audit—a framework for determining which recurring tasks should be delegated to AI based on five criteria: frequency/time investment, teachability, output verifiability, stakes of errors, and whether the user's involvement is essential.
Grok 4.6 Shows How Fast Your AI Options Are Expanding
Grok 4.6's release marks a significant shift in the AI landscape, with SpaceX AI re-entering frontier model competition alongside OpenAI and Anthropic. Meanwhile, the broader model ecosystem is expanding rapidly with competitive offerings from Chinese labs and open-weight models, while business adoption patterns reveal price sensitivity and enterprise concerns about data retention policies.
What the Heck is Graph Engineering?
The AI Daily Brief discusses graph engineering as a new concept in AI, distinguishing it from previous paradigms like prompt and loop engineering. The episode reviews OpenAI's model developments and a shift towards designing complex agentic systems for organizational tasks.
41 Stats That Tell the Story of AI Right Now
This episode presents 41 statistics about AI adoption and usage across enterprises, individuals, and society, revealing a widening gap between AI frontier users and laggards. Key findings show that 52% of US workers use AI on the job, but ROI remains elusive for most organizations, while emerging concerns around token costs and employee resistance are reshaping how companies approach AI implementation.