I have some coworkers that are similar in everything--education, work ethic, and intelligence--but some of the tick out ML ideas that work like clockwork, while others get hits rarely if ever. I cannot tell what makes it work for some and not others. Their ideas both sound equally good.
Sometimes a coworker will be an ML star for a year or two, but then suddenly run out of steam. It's brutal to watch.
I used to think most smart people had similar distributions of good ideas, and it was just that the hardest working tried out all 50 of their ideas to pick out the 2 good ones. But I've seen smart and hardworking people have a hit rate of 0.
I feel that the Zen used in the West and the Zen in East Asia are quite different. I think the Western Zen is probably the one from the 1970s book Zen and the Art of Motorcycle Maintenance. It usually carries a sense of equanimity and beginner's mind. But in East Asia, Zen actually emphasizes aimlessness or non‑purposefulness.
The point where I really feel the difference is that Western Zen seems to be about how to train the self to become stronger, whereas actual Seon (Zen) in East Asia is about going with nature, letting go of the self, and allowing things to flow. In the actual practice of Seon, it's about doubting the self, letting go of attachments, and realizing that achievement, comparison, and the desire for control are all just fleeting. There's a famous phrase: 'Banghasak (放下著)' — let it all go.
If anything, I think ancient Roman Stoicism feels more like Zen than Western Zen does
So that's fascinating. When I saw this article, I was expecting it to be about whether we should give up the desire for success, but instead it took a completely different direction, which was surprising
I don't think the intent of Robert Pirsig's work was to outline a git gud strategy cloaked in chillness. The book is heavily inspired by Eugen Herrigel's Zen in the Art of Archery which is explicitly not about trying to get good at something.
Both books highlight the value of dissolving conscious aim in favor of experience something. Pirsig's point isn't, you gotta act like a noob and then you can be good at maintaining motorcycles. His point was that there is a joy in losing yourself in these things that have to be done. If you are rushing to get it done or focusing too much on the end state, you will lose the joy and this thing will become a chore.
He does make the connection that years of doing things this way will lend you a kind of skill. And he connects the ideas to the Western concept of gumption which is a kind of motivation or persistence but again the book's core is not, lose yourself and you will get good. It's more like that a Western obsession with accomplishment can rob you of the joy that can come from engaging with activities for their own sake and not pushing through them just to get them done.
Zen I believe is about meditation where thoughts quiet down and you experience the moment. When you do, it means your brain gets rest because it is not "chasing thoughts". And that means when you stop the mediation you brain is well-rested and can work better because it can focus much better, the runaway thoughts are not leading it astray continually. That means your brain, and you, can accomplish more.
It is probably true that many Western zenists have figured this out and use meditation for that purpose. Think about star-wars, the "force" is about being able ro lift a rocket-sdhip with the power of your concentration. That's the myth.
Whereas I believe when you achieve a zen-like state of mind it becomes less important to you whether you can lift the rocket-ship or not. In the Eastern tradition Zen is it own goal.
Perhaps I've been deep in my own issues for too long, but it seems to me that the author is trying to say "don't trust the current evaluation suites too much"; scores only reflect a small part of the problem. What's interesting is discovering a new, stable evaluation metric, doing something new based on it, and having that new thing yield some unexpected intelligent results
This is certainly part of it! My point was that focusing on problems proposed by others is one very specific and pretty short-term mode of thinking. Good researchers improve benchmark scores. Great researchers think about what problem they're solving.
Around 2015, I found myself managing back end and machine learning engineers (not researchers) at the same time. Many of the back end engineers wanted to do more ML. Some of them did well when given a chance, but others wanted to revert to back end within a few months. At the same time, one of the ML leaders wanted to step away from ML and only do back end work to support ML.
As I studied these dynamics, something occurred to me... Different people need to see signs of success at different frequencies. Because of the nature of our product, measuring the performance of a new/updated model required the model to be live for at least a full calendar month. So, between initial work and final analysis, it was often a 2 month wait or more. For many back end tasks, you can build a quick prototype, run it to see if it works, and be on your way - the signals come all day long. The varying frequency needs of different people went a long way to determining which of them liked working on ML.
This is sort of a manager's version of feature engineering. ;-) The people on that team taught me a lot!
It revolves around the sentiment of "go deeper" - but I think it is a double-edged sword.
Sure, entropy, tensors and gradients are important - and yes, they are pretty much requirements.
But from what I see, it is the opposite - a lot (if not virtually all) progress in the last decade of deep learning was not because of a fundamental idea, but incremental, experimentally-verified practice.
Even though I think there is good intuition for why ReLU is better than sigmoid (tl;dr: last layer is log(sigmoid) ~ ReLU, putting anything different inside kills the gradient), the original paper by Hinton himself was more or less "because it trains 3x faster".
Re-thinking fundamentals might help, but most "let's change the fundamentals" is rarely how it works. Even the most seminal papers, i.e. AlexNet and "Attention Is All You Need", are refinements of existing ideas, and show how they help.
Machine learning is an experimental science. Many mathematically cool ideas do not work. Many engineering ones do.
> I've tweeted before that one of the most important traits in a researcher is healthy paranoia. Be paranoid!
I have seen so many PhDs burned out to cinders; I don't think it is any more a good piece of advice than "depression is good for philosophers". Sure, be a relentless explorer.
> In short, holding on to ideas for too long can actually be counterproductive. Stay open-minded and refuse to let ego cloud your judgement.
> If you want to solve a problem, the tried-and-true path to success is to attempt a solution, try it, reach a bottleneck, try to solve it, and only reach for literature when you’ve run out of ideas yourself.
I've found this to be the right balance between using your creativity and getting stuck too long
I think this also stems from ML being more like biology or alchemy and less like math or programming (where you can get down to the first principles, abstractions are rock solid, and non-determinism is limited in scope).
like the author said, so much of 'success' or 'progress' (in research but of course also across disciplines) depends upon temperament. just straight up having a good attitude about things. the skills that make a good researcher could not be more transferable: patience, innate curiosity, and a resilience against failure.
that said, these skills are increasingly rare/at a premium given our culture of minimizing discomfort tolerance via hyperconvenience. people have a harder and harder time waiting or failing.
Stepping away from the work to find inspiration, to allow the subconscious time to process everything, to present your conscious mind ideas is necessary. I try to pick a wild or almost outlandish idea from time to time, because if I only try what I think will work, then I'm not doing my job.
This reminds me of Ed Witten (greatest living physicist?) in an interview by Brian Green. Green asked Witten what his day-to-day was like at the Institute for Advanced Study ...
Haha. Unfortunately is my regular voice, since long before I started using Codex. You can check through some of my old writing. It definitely could've gotten worse though. Not sure if I'm training on Codex, or Codex is training on me...
> One really impressive thing about OpenAI is that most of the people running the company (on the technical side, at least) are under 35. Many of the important decisionmakers behind chatGPT are under 30.
During Gold Rush most 49ers were under 25, so there's still room for improvement!
[Continuing the analogy, you may find that many AI heroes are just those who happened to be closer to pools of TPUs and GPUs in the early days...]
On a similar note, I think it's pretty hilarious (short sighted) of Anthropic to have open hiring positions, but ban the use of it's product for... frontier model research.
Where do you think these people are going to come from?
Rash decision, and will likely draw an anti-competitive lawsuit at some point.
Tangential tidbit about etymology of the word Zen:
Zen is a Japanese word that comes from the Chinese “Chan”, which in turn comes from the Sanskrit word “Dhyana”, which roughly translates to focus/meditatiin.
That trajectory Sanskrit —> Chinese —> Japanese reflects the geographical trajectory of the spread of Buddhism out of India.
Same word in Vietnamese and Korean is “Thien” and “Seon” respectively.
26 comments
[ 2.7 ms ] story [ 44.6 ms ] threadSometimes a coworker will be an ML star for a year or two, but then suddenly run out of steam. It's brutal to watch.
I used to think most smart people had similar distributions of good ideas, and it was just that the hardest working tried out all 50 of their ideas to pick out the 2 good ones. But I've seen smart and hardworking people have a hit rate of 0.
The point where I really feel the difference is that Western Zen seems to be about how to train the self to become stronger, whereas actual Seon (Zen) in East Asia is about going with nature, letting go of the self, and allowing things to flow. In the actual practice of Seon, it's about doubting the self, letting go of attachments, and realizing that achievement, comparison, and the desire for control are all just fleeting. There's a famous phrase: 'Banghasak (放下著)' — let it all go.
If anything, I think ancient Roman Stoicism feels more like Zen than Western Zen does
So that's fascinating. When I saw this article, I was expecting it to be about whether we should give up the desire for success, but instead it took a completely different direction, which was surprising
Both books highlight the value of dissolving conscious aim in favor of experience something. Pirsig's point isn't, you gotta act like a noob and then you can be good at maintaining motorcycles. His point was that there is a joy in losing yourself in these things that have to be done. If you are rushing to get it done or focusing too much on the end state, you will lose the joy and this thing will become a chore.
He does make the connection that years of doing things this way will lend you a kind of skill. And he connects the ideas to the Western concept of gumption which is a kind of motivation or persistence but again the book's core is not, lose yourself and you will get good. It's more like that a Western obsession with accomplishment can rob you of the joy that can come from engaging with activities for their own sake and not pushing through them just to get them done.
https://en.wikipedia.org/wiki/Dhyana_in_Hinduism
It is probably true that many Western zenists have figured this out and use meditation for that purpose. Think about star-wars, the "force" is about being able ro lift a rocket-sdhip with the power of your concentration. That's the myth.
Whereas I believe when you achieve a zen-like state of mind it becomes less important to you whether you can lift the rocket-ship or not. In the Eastern tradition Zen is it own goal.
Is this something like what you meant?
As I studied these dynamics, something occurred to me... Different people need to see signs of success at different frequencies. Because of the nature of our product, measuring the performance of a new/updated model required the model to be live for at least a full calendar month. So, between initial work and final analysis, it was often a 2 month wait or more. For many back end tasks, you can build a quick prototype, run it to see if it works, and be on your way - the signals come all day long. The varying frequency needs of different people went a long way to determining which of them liked working on ML.
This is sort of a manager's version of feature engineering. ;-) The people on that team taught me a lot!
But from what I see, it is the opposite - a lot (if not virtually all) progress in the last decade of deep learning was not because of a fundamental idea, but incremental, experimentally-verified practice. Even though I think there is good intuition for why ReLU is better than sigmoid (tl;dr: last layer is log(sigmoid) ~ ReLU, putting anything different inside kills the gradient), the original paper by Hinton himself was more or less "because it trains 3x faster".
Re-thinking fundamentals might help, but most "let's change the fundamentals" is rarely how it works. Even the most seminal papers, i.e. AlexNet and "Attention Is All You Need", are refinements of existing ideas, and show how they help.
Machine learning is an experimental science. Many mathematically cool ideas do not work. Many engineering ones do.
> I've tweeted before that one of the most important traits in a researcher is healthy paranoia. Be paranoid!
I have seen so many PhDs burned out to cinders; I don't think it is any more a good piece of advice than "depression is good for philosophers". Sure, be a relentless explorer.
> In short, holding on to ideas for too long can actually be counterproductive. Stay open-minded and refuse to let ego cloud your judgement.
Which I think is true.
I've found this to be the right balance between using your creativity and getting stuck too long
like the author said, so much of 'success' or 'progress' (in research but of course also across disciplines) depends upon temperament. just straight up having a good attitude about things. the skills that make a good researcher could not be more transferable: patience, innate curiosity, and a resilience against failure.
that said, these skills are increasingly rare/at a premium given our culture of minimizing discomfort tolerance via hyperconvenience. people have a harder and harder time waiting or failing.
> on days we do not find insight, we sit.
This reminds me of Ed Witten (greatest living physicist?) in an interview by Brian Green. Green asked Witten what his day-to-day was like at the Institute for Advanced Study ...
Wittens' reply: "I sit at my desk".
I'm getting tired of the monoculture where both articles I read and agents I code with speak with the same (annoying) voice.
During Gold Rush most 49ers were under 25, so there's still room for improvement!
[Continuing the analogy, you may find that many AI heroes are just those who happened to be closer to pools of TPUs and GPUs in the early days...]
Where do you think these people are going to come from?
Rash decision, and will likely draw an anti-competitive lawsuit at some point.
Zen is a Japanese word that comes from the Chinese “Chan”, which in turn comes from the Sanskrit word “Dhyana”, which roughly translates to focus/meditatiin.
That trajectory Sanskrit —> Chinese —> Japanese reflects the geographical trajectory of the spread of Buddhism out of India.
Same word in Vietnamese and Korean is “Thien” and “Seon” respectively.