AI labs' power struggles shape enterprise adoption risks and rewards
🎧 PodShort
89 min squeezed to 2
AI SprinklerAS AI / ML
Paul Raiter
Founder & CEO at SmarterX and Marketing AI Institute
Mike Kaput
Chief Content Officer at SmarterX
Full episode from The Artificial Intelligence Show
Quotable Moments
It is just this like completely wild, unknown world we're heading into, where basically these five companies are going to decide everything when it comes to the economy, business, geopolitics.
But the way that the Wall Street Journal tells this is a lot of this does stem from Dario Amodei not getting the kind of credit he thought he deserved for his contributions to really the whole transformational phase we find ourselves in with language models.
You cannot make plans based on your current experience with these models. There is always a more powerful model in training. The labs have always already seen 6 to 12 months ahead of what you know to be true about reality.
Key Insights
- The conflict between OpenAI and Anthropic, stemming from personal wounds and power struggles, is profoundly shaping the future of AI development.
- Five major AI companies are expected to eventually control or influence everything in the economy, business, and geopolitics.
- The primary concern for users regarding AI is hallucinations (unreliability), not job displacement, according to a poll discussed in the episode.
- The AI industry is entering a phase where AI agents will automate AI researchers, potentially leading to recursive self-improvement and a rapid 'takeoff' moment for AI.
- The AI industry currently has three Tier 1 labs (Google DeepMind, OpenAI, Anthropic), two Tier 2 labs (Meta, XAI), and Microsoft as a potential Tier 3 player, with significant internal and external conflicts.
- Anthropic's leaked 'Claude Mythos' model is a significant leap in capability, especially in cybersecurity, and foreshadows an upcoming wave of powerful models.
- OpenAI has completed pre-training its next major model, codenamed 'Spud,' which is expected to 'really accelerate the economy' within weeks.
- AI forward managers with deep understanding of AI capabilities, domain expertise, and institutional knowledge are expected to thrive in the evolving AI landscape.
Metrics Mentioned
- 25% (Percentage of poll respondents who think OpenAI's enterprise deployment arm is a smart move.)
- 26% (Percentage of poll respondents with no opinion on OpenAI's enterprise deployment arm.)
- 28% (Percentage of poll respondents who believe OpenAI's enterprise deployment arm is inevitable.)
- 20% (Percentage of poll respondents who find OpenAI's enterprise deployment arm concerning.)
- 43% (Percentage of poll respondents who believe job displacement is still their top concern regarding AI.)
- 34% (Percentage of poll respondents who agree that unreliability is the biggest barrier to trusting AI at work.)
- 13% (Percentage of poll respondents whose biggest concern is 'something else entirely'.)
- 9% (Percentage of poll respondents who are not particularly worried about AI risks.)
- 45% (Percentage of organizations that officially approve 1-2 AI tools for employee use.)
- 34% (Percentage of organizations that officially approve 3-5 AI tools for employee use.)
- 15% (Percentage of organizations that officially approve 6 or more AI tools for employee use.)
- 5% (Percentage of organizations that block AI or have not addressed it.)
- 81,000 (Number of people in Anthropic's study regarding AI fears.)
- $1 million per day (Burn rate of OpenAI's Sora video app.)
- 500,000 (Number of users for OpenAI's Sora video app before it was shut down.)
- Hundreds of billions of dollars (Valuation of both OpenAI and Anthropic.)
- $25 million (Greg Brockman's political donation to a Pro-Trump Super PAC.)
- 6% (Drop in cybersecurity stocks (CrowdStrike, Palo Alto Networks, Zscaler, SentinelOne) after Anthropic's leak.)
- 7% (Drop in cybersecurity stocks (Okta, NetScope) after Anthropic's leak.)
- 9% (Drop in cybersecurity stock (Tenable) after Anthropic's leak.)
- 5,600 (Number of staff cut by PwC last year.)
- 70-80% (Uber CEO's estimation of human jobs AI will eventually replace.)
- 10-20 years (Uber CEO's timeline for physical jobs like driving to be replaced by AI.)
- $9 trillion (Meta's market cap target by 2031 to fully pay out its executive incentive program.)
- 500% (Required market cap increase for Meta to hit $9 trillion by 2031.)
- 97 million (Monthly downloads for Light LMM (open-source LLM).)
- 500+ hours (Amount of fMRI recordings used to train Meta's Tribe v2 brain encoder.)
- 700+ (Number of people from whom fMRI recordings were taken for Meta's Tribe v2.)
- $200 million (Amount raised by Harvey (AI platform for legal work) in its latest funding round.)
- $11 billion (Valuation of Harvey after its latest funding round.)
- $1 billion (Total funding raised by Harvey to date.)
- $1 billion (Minimum amount OpenAI Foundation will invest in 2026.)
- 26% (Equity stake OpenAI Foundation received in OpenAI.)
- $130 billion (Approximate value of OpenAI Foundation's equity stake in OpenAI.)
- 2029 (Google's deadline for migrating its systems to post-quantum cryptography.)
- June (SpaceX's target month for a public IPO listing.)
- $75 billion (Projected amount SpaceX could raise in its IPO.)
- $1.5 trillion (Last valuation of SpaceX.)
- 10,000s (Number of employees affected by Microsoft's hiring freeze.)
- 30% (Percentage of time sales reps spend actually selling.)
- 3,000 (Number of assets accidentally exposed by Anthropic related to Claude Mythos.)
RevBots.ai View:
- AI Sprinkler orgs face tool sprawl: 45% approve 1-2 tools but lack orchestration.
- Claude Mythos and Spud releases show ARM-stage labs outpacing enterprise readiness.
- Cybersecurity impacts prove AI adoption isn't just SaaS Hoarder tool decisions.
- Recursive AI automation will force ARM adoption faster than most orgs anticipate.
Join The RevBots ARMy
The insider daily for Autonomous Revenue Masters.