fooker

This was the fastest supercomputer in 1995, approximately the capabilities of a modern phone.

https://en.wikipedia.org/wiki/Cray_T3E

And if this was the fastest, the average ones were significantly smaller.

What you're missing is that the things you mentioned sometimes took months to run in the fastest computers thirty years ago, but you are mentally expecting them to work quicker now.

~1 bit models are pretty much decision trees in disguise! :)

Which LLM use cases can be solved with compact decision trees is an awesome open problem.

> Numerical high performance computing?

Any numerical high performance computing from 1995 will run on a 100$ android phone.

> Get rich following the bot's trade advice and never tell anyone about your bot.

The fun thing about open markets is the market will punish you for this with second order effects and it's in your best interest to get some whales on the bandwagon.

It could be untrue, but it's not obviously untrue.

The open weight models right now are nowhere close to ..say.. Fable 5.1

They could catch up if Anthropic reaches some sort of a plateau and they manage to keep improving, but there's no reason to be certain about that happening.

Whatever you can buy for inflation-adjusted-45$ in ~2035, almost surely.

You're assuming costs remain high.

Pretty much every technology in history is available at several orders of magnitude lower cost than the first few years of it being invented.

AKA eventually everything runs on a potato.

The sentiment they are trying extraordinarily hard to create is that in the long run Anthropic will be the only one with any AI worth anything.

That makes any loss or risk reasonable.

It's great propaganda, so great that they have themselves started believing it with certainty. This is one of the self owns that brought down the USSR :)

Of course they could be correct too, time will tell.

Communicating results when zero other people are working on the same problem as you and your advisor aiming to submit to a peer reviewed journal that takes six months to respond; versus when competing teams of hundred engineers working with a few billion dollars of funding trying to get the results before the other teams has to be completely different.

It’s astounding that people expect the later to look like the former.

It’s amazing for me to see math research going from craft to technology.

CS had this moment decades ago, sans some theoretical parts that are joining the bandwagon now.

I did a bit of math research in grad school collaborating with my math postdoc roommate. I found the aversion to any computation making its way into the craft to be odd decades ago.

Well, now the proof’s in the AI generated pudding.

> put an encryption key of 4

Hey the 4 was picked for good reason!

xkcd.com/221/

“PHP security team”

Everyone in such a ragtag team would have an overlapping day job where the same crisis management needs to happen.

That is the integral

When this piece of code exists in a library, it makes it pretty much impossible to extend by the user of the library.

It is important to get people excited about science, and inspire the next generation of scientists rather than choose the cheapest path to ‘good science’.

Without public support, you’ll soon find yourself without any funding. This pretty much explains NASA’s dire funding situation over the last thirty or so years.

You can only get so many people excited about weather balloons and space rock chemistry.

The country is pretty polarized now, and amusingly everyone seems to think that the polarization is one sided. Definitely not the side they support.

I have lived in many cities across three countries, and the price of chicken has always been a good proxy for the cost of living. And maintains the same ratio to the price of a bus ticket (and similar things).

Does this phenomenon have a name?

> Is the implication of all this that mathematics research should get a cash injection to pay for high-powered computational tools?

Math will, sure. Mathematicians, doubtful. After all, most high end physics research with billions in funding is now done by postdocs making a lower middle class salary.

Maybe partially.

I think this sort of progress, compounded over the next N years will serve to cleanly split useful math from the ego driven math you speak of.

Certainly there are enormously impactful pure math results yet to be found, but we have no idea really which directions to pursue.

I wonder if this is going to look like the chemistry-alchemy split in the rear view mirror.

“I’m on a war footing,” he said while grabbing coffee during a two-day math workshop at the University of California, Los Angeles, that was rife with talk about the existential crisis that artificial intelligence has created in his field.

There is a huge jump from

AI solves math problems to

> If there are no humans in the field

Realistically, math will be done by well funded teams of people instead of cult of personality mathematicians.

Also, insights from scientific discoveries are for everyone. You do not have to understand the complex multi dimensional algebraic geometry of general relativity to appreciate that gravity bends time as well as space. You don't need to understand fifty years of linear programming research to use a modern solver.

Math doesn’t exist to make mathematicians famous.

It provides deep insights into logic, computation, provides tools for us to understand the universe, establishes mind blowing relationships between seemingly unrelated things.

For example “how much information can a black hole contain?”

This phenomenon has a famous name, the clever Hans effect.

They will be, that’s the point.

The point is to make the field progress by discovering new things!

Who gets credit for what and how credibility is assigned is going to be muddy for a while but that’s the case whenever a bunch of new science happens in close sequence. History is full of math/science discoveries having debated and dubious provenance.

> who's opinion of Tao don't matter in the slightest.

Similarly, Tao being ‘on the warpath’ does not matter in the slightest when it comes to formalizing math and finding proofs to important unsolved problems with AI.

I think ‘destroy his field’ is ‘destroy his method of being famous in the field’.

Terrence Tao is famous because of solving problems better that most mathematicians because he has an uncanny ability to look across different math fields. AI gets everyone this ability for 20 to 200$ per month, eventually free.

Most mathematicians are happy to see interesting problems in their field being solved, and are looking to build on these to find more interesting problems.

No, pretty much every math PhD student and researcher is now using the AI+lean combo.

Depends on how you want to define fields.

Fields that seem really close to the layman are surprisingly difficult to have any knowledge transfer.

For example Programming Languages and Compilers.

What you are describing is a side effect of pursuing a PhD, not the main goal.

You can achieve the same side effects, often better, by pairing a smart undergrad with a good mentor.

Deep expertise is not about the soft skills here.