The AI lab watch
Tracking OpenAI, Anthropic, Google DeepMind + 2 more
- OpenAI — Frontier lab; track model, pricing and org changes as they are discussed.
- Anthropic — Frontier lab; track safety positioning, enterprise strategy and funding.
- Google DeepMind — Frontier lab inside Google; track research direction and product integration.
- Meta AI — Open-weights strategy and how rivals respond to it.
- Mistral — European frontier lab; track open-weights releases and enterprise deals.
OpenAI and Anthropic pivot to aggressive agent-based commercialization
~61 hrs of podcasts, listened to for you. This is the 2-minute version.
246 search results → 80 episodes scanned → 19 made it in · 15 analyzed in full
The frontier labs are pivoting from general-purpose models to autonomous agentic infrastructure, with both firms facing significant internal pushback on safety. As you track the shift toward token-based pricing and forced model shutdowns, it is clear that the industry is abandoning previous safety-pacing agreements in favor of a high-stakes, market-driven deployment race.
OpenAI
7 episodes · 11 of 19 mentions
Who Wins The AI Assistant Wars? + Why Didn’t Google Build Muse?OpenAI targets enterprise with $100/month 'Dots' tier · Big Technology Podcast · Listen
OpenAI is positioning 'Dots' as a premium enterprise tool, highlighting the fundamental divergence in worldviews between OpenAI and Anthropic.
Meet Dot: OpenAI's Always On AgentOpenAI's Dot integrates with 4,000 plugins · AppStories · Listen
Dot features persistent memory and local file access, marking a shift toward agentic models that execute tasks across local machine environments.
Googlebooks Have a Tiny Intel Problem - DTNS 5367Sam Altman defends accepting negative externalities · Daily Tech News Show · Listen
Sam Altman publicly stated the world should accept some 'bad outcomes' to realize the broader benefits of AI technology.
Why Did Another OpenAI Safety Insider Quit?Safety systems lead resigns from OpenAI · Tech Brew Ride Home · Listen
David Robinson resigned from OpenAI, citing that the company's current 'trial and error' culture is insufficient for the risks posed by modern AI capabilities.
How OpenAI uses ChatGPT Sites (live at DevDay!) | Kath Korevec (Product Lead)OpenAI launches 'Sites' for software deployment · How I AI · Listen
The new 'Sites' tool allows users to build and deploy software using Codex, integrating with 60 enterprise connectors like Slack and Snowflake.
Nike in Freefall, OpenAI Fires Spies, & G7 Dumps 100M Barrels | PBD #867OpenAI fires safety researchers over confidentiality claims · PBD Podcast · Listen
OpenAI terminated three safety researchers, fueling speculation that the move is designed to clear internal hurdles ahead of a potential IPO.
TWiT 1104: Real Gs Move in Silence Like Lasagna - An AI Data Center in Space?OpenAI shifts to token-based pricing for enterprise users · This Week in Tech (Audio) · Listen
OpenAI is moving away from flat-rate subscriptions toward token-based billing and is acquiring thousands of Mac Studio M5 Ultra units for internal training.
Anthropic
5 episodes · 10 of 20 mentions
Oct 5 · 74 min · Top 25 business
Anthropic faces FTC inquiry and extreme revenue concentration
- The FTC opened an investigation into Anthropic and OpenAI regarding potential safety and consumer protection concerns.
- Reuters reported Anthropic's S1 filing shows the company incurred an operating loss of $8 billion last year.
- Data indicates 80% of Anthropic's enterprise revenue is currently concentrated within just 1% of their customers.
Why this matters to you
Extreme customer concentration and regulatory scrutiny create significant downside risk for Anthropic's upcoming public offering and long-term valuation.
“The most frightening thing about this is I don't think you've ever seen a company ask for 500 billion in market capitalization that had a fraction of the client and revenue concentration that Anthropic had.”
— Scott Galloway
Listen Deep dive
Also mentioned
Has the Memory Cycle Broken? Plus: Nvidia & Broadcom’s $40B+ Financing PlaybookBroadcom to loan Anthropic $42 billion for TPU construction · The Circuit · Listen · Watch
Broadcom is financing $42 billion for Anthropic to build Google-designed TPUs, signaling a trend of vendor-led balance sheet support for frontier labs.
20VC: Is Seed Investing Dead Without a $1BN Fund? | Does Ownership and Price Matter When Companies Can Be $1TRN Exits | Are AI Revenue Numbers Real and What to Watch Out For with Venky Ganesan, Menlo VenturesMenlo Ventures concentrates 20% of fund in Anthropic · The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch · Listen
Menlo Ventures secured a position in Anthropic by committing 20% of its fund, prioritizing access over high ownership stakes.
Facing Our Investing FearsAnthropic S1 filing prioritizes existential risk disclosures · Motley Fool Money · Listen
Anthropic dedicated 80 pages of its S1 filing to potential existential risks, nearly double the space allocated to explaining its core business operations.
S&P 500 Edges Closer to Record, Shrugging Off Bond & Oil RisksFormer Anthropic researcher dismisses AI 'kill switches' · Bloomberg Businessweek · Listen
Former researcher Jacob Coxon testified that proposed AI kill switches are merely 'tinkering around the margins' of long-term safety.
Meta AI
1 episode · 2 of 5 mentions
Putting Meta’s New AI Agent to the TestMeta launches 'Muse' consumer agent · WSJ Tech News Briefing · Listen
Meta’s new 'Muse' agent handles shopping and email, positioning itself against OpenAI’s Dots with a focus on privacy-encrypted virtual environments.
Make it yours
Start from these trackers
Copy its trackers, shows and tone into a page of your own. Change anything — this page isn’t touched.
Make it mine