May 29 2026
- Bill Gurley joins the show!
- Becoming AI Native
- 2 kinds of people - Using AI to learn faster and more than you ever could and those who use AI to avoid learning
- Proficiency in Claude is the most single most marketable skill right now (maybe just a short term arbitrage)
- You can even have a dialog about a prompt with Claude
- Pope releases encyclical on AI
- Warns business leaders to protect humans from AI
- Anthropic co founder Chris Olah joined
- Question is how much power do you give the government
- Leo the 13th warned about the industrial revolution - he got the whole thing dead wrong clearly
- Work week 60hr → 34 hrs globally
- real wages 8-10x adjusted for inflation
- median worker now earns more than a doctor did in 1891
- global GDP per capita went from $1500 → $20K
- Child labor in US went from 18%→ 0
- Work place deaths fell by 40x
- Life expectancy went up 60%
- global poverty went from 75% of humanity to < 10%
- All this happened because of technology, innovation and capitalism
- Anthropic enigma
-
Leading the field and also most negative about their field
-
They are aggressive lobbyists
-
Bill Gurley has a new theory
- Dr Frankenstein theory
- Read up
- Claude has a constitution
- Chris Olah
- Amanda Askell podcast appearances
- Dario writing (machines of loving grace)
- See how it ends (are they asking for overlords?)
- Also postulates about an AI that will reward humans
- Seems like they think they are mid-wifing a deity here
- Seems like they want to create a god
- Their tactics are also game theory optimal
- Absorb more capital
- Create a regulatory body where the refs don’t understand the game and you will run over the game
- They have a halo with their doomerism and create empathy
- Their rhetoric that they are the ones who care about safety and others don’t and AI recklessness needs to be controlled then you can see their strategy is smart
- Regulatory capture follows
-
If AI is so powerful it should be decentralized else one company beholden to the government will dominate humanity
-
Apple might be the darkhorse here if open source takes off and you can run AI on your own hardware and get AI sovereignty
-
And it seems there are multiple powerful models - so these AIs are getting commodotized? So evals are continuing to asymptote
-
Industry needs models to be swappable - so maybe the need is for an open source harness
-
Seems some companies are cutting back on their AI budgets half way into the year - even MSFT cut back on their Claude licenses
-
Sacks warning about rhetoric from Anthropic and folks in Washington how they keep warning against open models wrt guardrails not being available
-
Elon claims SpaceX has trained an in house AI for an order of magnitude cheaper (using C rewrite)
-
Bill Gurley new blog - from open source software to open source strategy - P3 institute
- open source is the backstop
-
Seems Anthropic is pulling ahead of OpenAI
-
Recent story about how some company spent $500M on Anthropic/Claude in a month (to be confirmed/identified if true)
-
- AI Jobs Debate
-
Sam Altman and Dario Amodei walk back their AI jobs apocalypse rhetoric
-
Goldman Sachs CEO also warns against AI job apocalypse (in NYT article)
- Is he auditioning for their IPOs? 🙂
-
So Sacks has been proven right - same as Jensen
-
There was no evidence for what the doomers were claiming
-
Code is now the #1 use case for AI but software hiring has not gone down due to AI
-
Software engineers are being hired more
-
Github has 10x more code now so you need more engineers to manage that code
-
You can now hire software engineers for firms that would never have hired them before
-
Job postings for software engineers are up 15% YoY in spite of coding being automated largely so AI is clearly creating jobs
- Not counting all the infra build out jobs being created
-
Clearly poor management by CEOs is getting the cover using AI as the scapegoat
-
The reason why the big companies over-hired was to keep the folks off the market to keep them away from startups or competitors
-
What AI doomerism misses is that competition exists
-
Comments