One advise I'll never forget (from an interview with late mathematician Maryam Mirzakhani) is that you don't have to race faster than everyone else (although if you can, that's great!). You can just run in a direction no one is running toward (and hopefully be persistent). Then you're only racing yourself and speed is largely irrelevant.

I myself know I'm very slow, specially developing projects. So I don't try to race anyone, except contribute to things I know are neglected. There are so many neglected problems, you don't need to work on the most coveted fields.

Problem is that's not how academic funding works. Grants are generally not given to possible dead ends.

Research grants are typically given with the expectation the project may possibly be a dead end. Otherwise it wouldn’t be research.

In theory, yes; in practice, no. Many grants these days are given based on the trend du jour.

Sure, but funding trends is different from funding things expected to work.

Funding things you wont be fired for funding. IBMs!

I’ve seen a lot of grants end with “we learned something”. Is that a dead end? IDK.

“Run in the other direction” is so generic that it isn’t helpful. No one else going in that direction is, at best, unrelated to the quality of the idea and, at worst, a heuristic that the idea is bad.

It's not the other direction, it's an other direction. The distinction is important.

If there are only two directions and everybody runs in one of them, there's probably a good reason. If there are virtually infinite directions, trusting your gut about an unexplored direction that you find interesting is a much better bet. See also: startup advice.

When asked how to keep up with all of this AI/ML stuff when starting your own AI business, Jeremy Howard (general genius and co-creator of fast.ai) replied that you don't need to know everything and keep up with everything. Pick a niche and focus just on that. And if that keeps changing too fast or is too much, pick a niche inside that niche.

Don't be discouraged. Your work is good.

And remember we are 6-12 months off having good, open source chatGPT class open source models and the software support to make it possible to run them at home.

We're also 6-12 months off from having the first strong AI restrictions.

Ironic mantra for you: Geoffrey Hinton in 70s, Minsky, Symbolic Logic, Neural Nets, stay the course.

I've almost given up just trying to keep up with the field... let alone competing. I'm not sure how people are doing it (keeping up) at this point given the pace of change.

don't give up max, you're tremendously inspirational to the rest of us who are even further behind in trying to keep up with what's happening. you're by far one of the best simplifiers/explainers of whats going on who actually regularly builds useful stuff. its just a matter of time until you hit the thing that goes vertical, and all this work to date will have been practice. you can only connect the dots looking back.

thanks :)

I've called this the rapidly shrinking innovation-disruption-adaption cycles. It quickly becomes the new rat race.

"It will be unlikely we will be able to plan for the disruption as it will simply happen before anyone can reason about what is likely to happen next or what to do about it. Any such planning efforts will be obsolete before they begin"

https://dakara.substack.com/p/ai-and-the-end-to-all-things

The problem is the dynamic of AI is leveraging massive computing power into something flashy - OpenAI and DeepMind have been going back and forth on this for a while, with OpenAI leverage a lot of human also in ChatGPT.

It's a bit like the original dot-com bubble, when people were throwing what was then vast amount of money out just to be in the space. And not knowing what the business model was going to be. The business model then turned out to be advertising but it's certain that advertising is well suited to a chatbot or other new AI stuff, stuff that is computationally and dollars-wise expensive.

AI projects are often valuable if given away. I assume people will ask how valuable per watt they are after a while.