CW

I’m on a mission to civilize. Progress is slow, but I’m in it for the long haul.

[CW @carlinw]

@ajobst @Mr_Derivatives How would you count things I previously searched about (vacations, professional research) that I now use a chat bot for? They do web searches as a part of their chain of thought but the revenue/cost per volume is very low compared to end user search?

[CW @carlinw]

@NilsHeadley @ajobst @bobbyfijan The original purpose of the test was to sort WW1 recruits. In any case, Andrew's point is that 30 years ago there was signal in the difference between a 1500-1600. That signal was helpful to the most selective 15 universities.

[CW @carlinw]

@ajobst @bobbyfijan I noticed this a couple years ago when I was mindlessly looking at college statistics. I was looking at the SAT scores for 25% and 75% percentile. By the 3rd school I realized these numbers must mean something different than they used to. 3 minutes of Googling later...

[CW @carlinw]

Rare PG comment I disagree with. Agree there is some genetic component.

Effort is domain-specific. Kids who drift in school often grind obsessively at gaming, music, or sports.

Carol Dweck's growth mindset work shows that whether a kid believes effort is worthwhile shapes whether they try. That's teachable.

Pro athletics effort changes considerably even well after they reach the pro level.

[CW @carlinw]

Agreed. I’m waiting to see whether a tool eventually comes in to existence that allows people to make shows like Taylor Sheridan or Shonda Rhimes but with a small team of 3-4. Landman Season 1 had a budget of $150m. If costs get down to $15k to $150k for a season, creators will have the ability to self fund or have a really amazing crowd sourced approaches.

[CW @carlinw]

Facts, Honesty, Opinions and Truth https://t.co/x9f8AaxW2L

[CW @carlinw]

There has been controlled selling for years. It’s how some people have accumulated pre-IPO shares. In transactions I have seen ROFR has only been exercised once.

Also - they have access to material non-public info such as the real timeline for AI satellite which appears to be sub-30 months.

[CW @carlinw]

Superintelligence Part 2 https://t.co/ZIlVo39MdN

[CW @carlinw]

This has aged extremely well. I like the local lens. Is there enough housing from Burlingame to San Jose? Not hard to do the math of if you’re making $300k at Apple can you afford a home? Average income worker like teachers you can see from commute durations and traffic. Here in Cleveland school teachers can afford to live 15-20 minutes away from their jobs.

[CW @carlinw]

@JosephPolitano Who is buying at this volume? Is this primarily bought by independent producers?

[CW @carlinw]

Leaving California... https://t.co/LxsAZhWEXD

[CW @carlinw]

Look, no hands! https://t.co/38PSwZYZgS

[CW @carlinw]

Writing publicly... https://t.co/GvtD4m6wP4

[CW @carlinw]

1 year later. Use FSD to drive 91% of my daily miles and typically 0 interventions and definition none to avoid accidents. Huge progress in the last year!

[CW @carlinw]

I have had 2 Teslas with FSD. I have grown to completely trust FSD to drive both city and highway in CA supervised. Would often drive from Sacramento to SF at night or in rush hour with 0 interventions in 88 miles. Have since moved to Cleveland Ohio about 2 months ago. Have had to intervene to avoid dozens accidents, some very serious (round about, freeway exits, express lane confusion). I had 4 today. I am a huge Tesla bull both as customer and investor. But let me tell you this stuff is NOT the same experience everywhere. I would not buy FSD at current quality in Ohio. I am confident it will get better. @dylanloomis

[CW @carlinw]

I think they can fix short term scaling quicker if they wanted. It’s likely a mix of operational and confidence/risk management. I do think both will win for some time (3-5 years). I’m long both stocks. I drive about 15k miles a year with FSD. But intuition is, I’d take quicker expansion from Waymo over next let’s say 2 years than Waymo out executing in 2-4 years. Waymo is very obviously the leader atm.

[CW @carlinw]

Opinion | Harvard Derangement Syndrome - The New York Times https://t.co/ajM76qxXHY

[Florence Bourgeois @ftbourgeois]

See also @statnews: https://t.co/N9jf7x6rMn

[Florence Bourgeois @ftbourgeois]

Our study in @JAMAPediatrics finds that few AI/ML-enabled devices have been developed for use in children and many lack critical data on validation in pediatric patients:
https://t.co/JnyqKEVdST

@HarvardMIT_CRS @RyanBrewsterMD