# Anthropic CEO Dario Amodei predicts a Nobel Prize for cofounder Chris Ohla 

*CONTEXT🎸🤖JAMMING* · https://www.youtube.com/watch?v=ZURFyhoyH20

## Summary

Dario Amodei discusses interpretability as a key to steering and making safe AI systems, predicting that his cofounder Chris Ola will win a Nobel Prize in Medicine for insights into mental illness. He also recounts his own path into AI research, including a six-month self-study plan after a Twitch contract, and how he messaged Greg at OpenAI to get a job there.

## Key points

- Amodei says interpretability is both the key to steering and making safe AI systems.
- He predicts Chris Ola will become a Nobel Medicine Laureate because interpretability may explain mental illnesses like schizophrenia and mood disorders.
- He argues neural nets are easier to study than brains and can be learned about in three to six months.
- Before entering AI, he did a three-month contract with Twitch to fund six months of self-study.
- His 2015 self-study plan included a Coursera machine learning course, Kaggle projects, Linear Algebra Done Right, and a statistics textbook.
- He bought a GPU with YC alumni credits and SSH'd into it to work through courses.
- He messaged Greg as soon as OpenAI was announced, offering to help despite getting a B minus (or C plus) in linear algebra.

## Notes

### Interpretability and AI safety
- Dario Amodei calls interpretability the key to steering and making safe AI systems.
- He says interpretability contains insights about intelligent optimization problems and how the human brain works.
- He predicts, seriously and not joking, that Chris Ola will be a future Nobel Medicine Laureate.

### Why neural nets may explain mental illness
- Amodei was formerly a neuroscientist.
- He suspects mental illnesses that remain unexplained, such as schizophrenia and mood disorders, involve some higher-level system phenomenon.
- Brains are "mushy" and hard to open up and interact with, while neural nets are not a perfect analogy but will become a better one over time.
- In AI, you can learn a lot about how systems work in three to six months.

### His six months of self-study
- After running out of personal runway, he did a three-month contract with Twitch to fund six months of study.
- His 2015 plan: a Coursera machine learning course, Kaggle projects, reading Linear Algebra Done Right, and a statistics textbook.
- He used YC alumni credits to buy a GPU and SSH'd into it to work through courses.
- This was after AlexNet, so he mostly studied image classification.
- He notes this plan is probably not the right one for people today.

### Getting the OpenAI job
- At the time, DeepMind, Google Brain, and Mila were the three places he considered for AI work.
- He messaged Greg as soon as OpenAI was announced, saying he got a B minus (possibly C plus) in linear algebra but knew engineering and distributed systems.
- He offered to "mop floors" if they needed help.

## Transcript

[00:00] And so places where I think that's about to happen and it hasn't happened yet. Um, one of them is interpretability. I think interpretability is both the key to steering and making safe AI systems um and and we're about to understand and interpretability contains insights about intelligent optimization problems and about how the human brain works. Um, I've said and I I'm really not joking. Chris Ola is going to be a future Nobel Medicine Laurette. I'm serious. I'm serious because a lot of these I used to be a neuroscientist. A lot of these mental illnesses, the ones we haven't figured out, right? Schizophrenia or the mood disorders, I suspect there's some higher level system thing going on and

[00:42] that it's hard to make sense of those with brains because brains are so mushy and hard to hard to open up and interact with. Neural nets are not like this. They're not a perfect analogy, but as time goes on, they will be a better analogy. In AI, it's really something where like in three three to six months, you can really learn a lot about how these systems work. It's not not so complex. And I think uh I think it's probably worth learning if you think it's plausible that these systems are going to become that powerful because I think we need our best and brightest to kind of I don't know have their game face on or something for uh the challenges and the opportunities that that the technology might bring. where I

[01:18] had like my personal runway. I I ran out and so I was like, "Okay, I think that I'm going to need six months of stealth study to have a shot at getting a job." At that point, it was Deep Mind or Google Brain were the two places to do work there. Or Merie Merie was the third one that I was like looking at. So, I was like, if I want to help out with that, those are the three places to look at. I don't have any of the skills yet. I need six months of self-study to feel like I would not be a drag on them and like actually be helping instead. Can you um maybe explain a bit what was that cell study like because I'm sure there's a lot of software engineers right now in their 20s are looking tool to become AI

[01:56] researchers. What was what was that 6 months like even though as you said you had uh gotten a B minus in linear algebra which is like core >> might have been C plus I'm checking going to give >> that's pretty impressive where where you got to. >> Yeah. Yeah. It turned out okay. First I did a contract actually with Twitch um and like earned like enough to have that six months of runway. So I did like three month contract with Twitch and then I made a a plan to self-study. I don't think it's the right plan now for people too because at least the 2015. What did it look like? It was like take a Corsera course on machine learning, try to solve some Kaggle projects, read linear algebra done right and uh I had a

[02:36] statistics textbook. I think I had YC alumni credits and so I bought like a GPU and I would like SSH into the GPU to like work through my courses for it. >> And this is right after Yeah, it was already after Alex Knack, right? >> This after Alex Net. Yeah. So I was mostly doing image image classification stuff that I was trying to learn was like the thing that all the courses would teach you to do. >> How did you get the open AI job? Because you were one of the few engineers. it was mostly researchers and they had pretty stacked team of researchers. >> I messaged Greg um as soon as OpenAI was announced and I was like, I'd love to help out in some way. I got a B minus in my linear algebra, but I know some

[03:17] engineering. I've done a bit of distributed systems work. If you guys need help, I'm like happy to mop floors if if