Builder, Dad - YC S13 - Cofounder @PlayPokerSkill, @Cratejoy, MarketZero - Question the answers
Austin, TX
Joined December 2008
- Tweets6.9K
- Following816
- Followers1.3K
- Likes3.4K
Producers of the revolution: great in the window, poor over the long run. In 1900, railroads made up 63% of US stock market value, and nearly 80% of US market value that year sat in sectors that are now small or gone.
I'm sure this time its different.
isn't it weird how publicly traded companies with frontier models don't seem to have apocalyptic news to announce?
Amir Elaguizy retweeted
In 1994, software was dead. People told me programming was for nerds and that I should become a doctor or a lawyer instead.
In 2004, software was dead. People told me that someone from India would take my job for $5/hour.
In 2014, software was dead. People told me that no-code was going to kill every job and doom my career.
In 2024, software was dead with Devin, Windsurf, Copilot, and OpenAI o1.
We kill software every 10 years, and yet the demand for good Software Engineers has never been higher.
linus torvalds is a gift from heaven unto humanity. a grumpy angry sometimes slightly incoherent gift from heaven. his rants are typically totally right. always entertaining. generally offensive. exactly the leader open source needed.
Linus Torvalds to the anti-AI crowd: stop using compilers if you want to prove they’re REAL programmers. LOL.
“you had better not use an editor to write that commit message either.
The kind of expectations *I* have is that people do their best and I expect people to understand what they are doing.
But that understanding has very little to do with what tools they use.
And quite honestly, in my experience, the people I respect and call intelligent are not the ones that shy away from using new and interesting tools.”
dot is the only openai product i currently use. first time ive had no use for codex at all. it just doesnt matter right now, the models are behind and no random features will fix that.
if you do not work on your process for managing concurrent agents they will burn you out. you should not be the bus and you shouldn't let them concentrate decisions up to you.
Just assume that if you use Anthropic or OpenAI your accounts will eventually get banned and plan accordingly.
hard to overstate how much of a quality of live improvement it is to speak to lawyers 90% less as AI has improved
argon is pretty good actually. Its by far best and most usable model they have ever made. its noticeably slower than opus, and noticeably smarter than sol
The difference in behavior from publicly traded companies building models vs privately held startups is stark.
I forgot I had asked sol to do something simple. Realized it had been running for 2 days (41 hours). Turned a like 200 line feature into 7,000 lines of code mostly harnesses and tests and insane levels of documentation.
This is nuts.
Amir Elaguizy retweeted
It’s becoming clear that the notion of AI alignment “is” the safety problem in AI.
yeah or he could use AI to do more and more challenging math rather than cry that the his existing process has been levered up
“On Tuesday night, I cried, and it woke my daughter. [OpenAI] annihilated my field of research. They wanted to prove they could massacre us.”
these are the words of Fields Medalist, Hugo Duminil-Copin, who was working on 16 problems OpenAI now claims it has solved.
and he says some mathematicians may now keep their research problems to themselves rather than share them publicly.
AI might actually make mathematics less open.
the period of fantasy capitalism is almost over - pacing is dying - safety researchers are getting fired. bring on the acceleration!
i've noticed more and more of my friends who had psychological aversions to using terminals switching from AI apps to straight terminals over last few months. i think their trauma centered on the arcane bash commands learning curve and now they see terminal doesn't have to suck!
if i were to define super intelligence when i was in college 20 years ago what we have now would have cleared that bar but i would have been confounded that it both clears the bar and stops (eventually) if you dont keep typing and has perpetual amnesia. ??? mensa + memento?
this is insanely true if you've been deep tech for your entire life AI feels more like a progression from algorithms to machine learning to deep learning to real AI, kinda slow tbh
