I still think the fire metaphor for AI is underused:
"Is fire good or evil?" That's a weird question
"Should we give everyone access to fire?" Yes, so they can cook and heat their homes
"Should we give everyone flamethrowers?" Probably not
"Should we give everyone firebombs?" Definitely not
"Should we require companies that produce enormous combustion reactions to do so safely?" Obviously yes
“Should we wait until the city burns down to establish a fire department?” Obviously no, and we should also build in safeguards like smoke detectors
Make the margins next to zero for all these AI models. It was trained on humanity's data, it should be gift to ourselves. Doing so will save us from a few in control of our species.
Distill at industrial scale! Distill, I say!
Marc Andreessen: There are two ways to think about education. One is at the national level — how do you educate all kids? But the real question is N = 1: what do you do for one individual kid? And for centuries, the answer has been obvious.
If your goal is to maximize a single child, the best method by far is one-on-one tutoring. Every royal family knew this. Every aristocratic class knew this. It’s why Alexander the Great was tutored by Aristotle — and then took over the world.
There’s actually statistical proof of this. The Bloom’s 2-sigma effect shows one-on-one tutoring can move a kid from the 50th percentile to the 99th percentile. No other educational method comes close.
AI changes that. For the first time in history, every kid can have access to infinite questions, instant feedback, personalized explanations, and real-time quizzes — all at N = 1 scale.
This is the most powerful shift in education we’ve seen in centuries. One-on-one tutoring was always the gold standard. AI is what finally makes it available to everyone.
What I find endlessly fascinating:
Some engineers really can't seem to grasp that LLMs are non-deterministic and how to build software taking that into account.
For others it immediately clicked, but for some it seems like there's a real mental barrier to accept it.
Lots of folks are wondering why I'm suddenly making videos 😅 Thought I'd address that in today's video!
Also, tried a different format today: I'm not walking around, and the video is square. Do share feedback.
Going to keep iterating!
The thing that has always fascinated me most about LLMs is how incredible they are as products of compression rather than intelligence. A few gigs can store a bulk of the public internet - truly wild.
Google has the finest LLM,
Yet fewer come to seek its lore,
ChatGPT is used far more.
Why? Sundar lacks what Sam excels:
the art of hype and tweet.
The fate, it seems, was long decreed:
"Attention is all you need".
New Blog Post: Are we entering an age of abundance?
Technological progress is accelerating very rapidly and I don't think most people are prepared for that future.
My bet is that the AI coding agents quickly beat top 0.1% of human performance, at which point it wipes out the need for the majority of software engineers.
We will still have senior software engineers supervising the AI for many years. But they will eventually be like the conductors on self-driving trams; they are mainly there to make people feel better.
Software engineers are highly-paid farmers, tending their crops by hand.
We just invented the combine harvester.
The world is going have a lot more food and a lot fewer farmers in very short order.
@NWischoff@AlexReibman We're a newer shop (https://t.co/tAQM7hovAU) but have underwriting experience. Same deal applies if we build a better POC using AgentStack?
Also not raising :)