Hacker News Reader

Comment by theturtletalks | Hacker News Reader

I just watched Primeagen’s video on this and Tao’s point is that juniors no longer have the path of solving a proof to earn Field’s medals. He also argues that the community part is being hurt by AI discovering proofs because in the past people used to get invite to talk and collaborate. Now all that is being taken away. The community must adapt because Pandora’s box cannot be closed.

Prime also mentioned that software development is different. In Software development, the product is what you’re building towards, so the means to get there can be disrupted without the industry being cannibalized.

In math research, the process is the product. You take away the researching part and not much is left. But my question is, these math proofs OpenAI released, will math shift to actually using the proofs to change the world instead of just finding new ones?

View on HN

Our industry is equally cannibalised, anyone that thinks otherwise is either having too many tokens or in a privileged position.

If business can deliver the same product with a smaller team, great!

And yes this has been happening for a while, even if not everywhere.

In enterprise consulting, projects that would require a team of 20 devs on average, now have about 5.

Moving away from on-prem, managing own cloud infra to managed containers, to serverless, SaaS and iPaaS ready made products, and offshoring naturally.

All contributed to ever decreasing team sizes.

Now AI based tooling is added to that cocktail, reducing even further the team sizes.

The only folks doing well in the end, are the employees of AI companies, without moral issues contributing to the industry downfall, because the CEO themselves aren't the ones coding and pirating human culture.

For now, a skilled person using AI is still miles better than an autonomous AI building something. I’ve been trying to do the latter for months to build open source alternatives and the end products still lack polish and that last 20%. Maybe this changes, but I still think there will be people who can use that AI to be better than AI alone.

Definitely, the problem is that companies need less of them, just like a construction company opening roads with machinery, or an automated factory.

It could be that there are more companies though, companies that couldn't afford a large team before. Now they can start delivering value with just, say, 3 people, making the business worthwhile. At least that's one optimistic scenario.

For now that is balanced by the fact that we build more software, things that wouldn't be worth it before.

If they weren't worth before, they aren't suddenly worth paying for now.

If the cost comes down, that could make something worth paying for when it previously was not.

Not enough to keep that exponential growth curve going.

Wait, I thought we were discussing loss of employment of engineers? Whether the AI companies continue their exponential growth curve is another question entirely. Making up for the (thus far hypothetical) loss of engineering roles doesn’t require the AI companies to even stay in business, let alone grow exponentially.

It's not balanced, the market keeps shrinking visibly.

There's CVs out there that would've made recruiters go mad 36 months ago sending their hundreth application.

Number of employed SWEs has maintained the same, slightly increased in fact.

Which companies?

Doesn't matter, if the number of position keeps shrinking so will our job prospects.

I'm a freelancing consultant since 5 years, I've had 2 major customers now for 3+ years. I have a very good pulse of the market: being good, or being even very good and being among those that brings AI and automation to organizations will not save our jobs.

In fact, AI has sped up so much the work that 2 out of 5 people in my current team are being let go: I find it absurd, our productivity has more than doubled over the last years and we've made ourselves redundant. Money is money, I'm on one side making non-tech workers redundant (people whose job was menial boring office stuff), and building the systems that will make myself redundant.

An industry being able to deliver the same value to society with less human capital isn't a "downfall." It sucks for the people who used to be able to extract lots of value for themselves, but for society and the industry, it's a plus.

It is a plus for the shareholders, that is for sure.

For the society definitely not.

Do you believe that farming technology, freeing up 95% of humans to not have to toil in fields all day, was only beneficial to shareholders?

Thing is that people don't want to be "freed up" from their comfortable, high-paying jobs. Because then they likely need to return to metaphorical toiling in the field.

Read the second to last slide. What we need now is *imagination*. You assume the need for new software is fixed and that AI is going to satisfy that need with fewer humans (lower cost), but by lowering the cost we can increase the supply of software!

That means software engineers better start getting creative. If you think your job is to wait for a PM to assign you a well-written researched ticket, you're done. Your job is now to figure out how to make these machines (computers) do whatever we need them to do safely, quickly, at scale, and correctly by applying all your knowledge of computer science and the engineering field of software engineering to an AI prompt.

This doesn't scale, because like in a factory that gets replaced by robots, or in a supermarket with self checkouts, not everyone gets to save their job, regardless.

Also, the increase in output is meaningless when the amount of customers doesn't scale in similar size.

Then there are the constraints of physics, there are so many humans in the planet that actually want to pay for a specific product, or consulting services.

I read the "imagination" as prelude to "Jevon's paradox" for AI. It might be that we simply don't know yet what using AI fully means, and where else it can be used and how. Electricity was initially just for lighting, but now it's everywhere...

> Tao’s point is that juniors no longer have the path of solving a proof to earn Field’s medals.

This is a "you" problem for the math establishment, not a problem for the AI companies.

> This is a "you" problem for the math establishment, not a problem for the AI companies

This was a talk given to other mathematicians about the future of mathematics; sounds like only "you" have a problem for some reason.

I'm sorry if a factual statement makes you unhappy.

Entirely correct. The math establishment needs fundamentally overhaul its incentive structure-irretrievably broken-to function under the assumption that AI involvement in research is completely ubiquitous.

I'd go beyond that and say they need to overhaul their culture and mode of operation. Math needs to be even more collective than it is today, without focus on ego reward and priority. They were already steps in this direction before this year's AI detonation: net-enabled collaboration, first informally and then with Lean formalization. Perhaps there should be a de-emphasis on naming things after people.

Your first statement answers your question and the answer is no.

To use a given proof you have to understand it. Galois theory, long ago, began with the question "what is the formula for a sixth or higher polynomial" and ended with the answer "there is no formula involving only nth roots". The answer isn't that interesting but the methods of Galois involved showing a deep relationship between rings and groups and that is still interesting together.

OpenAI's proofs reputedly aren't very or at all comprehensible so these raw proofs aren't immediately useful. Maybe mathematicians will work them into useful and understandable to humans approaches. But that asking something from the humans, right?

Primeagen is a youtube reaction guy. You may as well tell me what hbomberguy, Jay Leno, or Ben Shapiro had to say about it.

I mean this is funny ad hominiem but OP has a point. Academia has always massively prioritised understanding over outcomes, injecting startup culture into it is basically injecting antimatter.

Thesis, antithesis, synthesis?

He also has an AI harness named after him (Prime Agent) so there's that.

Reply on news.ycombinator.com