Remix.run Logo
▲ quasigod 7 hours ago

This title seems weird. Why just "anthropic employee", not specifying that hes a mathematician, or just including his name? Without the clarifications in the replies, many will likely assume this was a random Anthropic employee just throwing tokens at the problem, not a very well known mathematician working with the model. Probably not the intention, but these kinds of titles feel like they're trying to shift the credit from the person to the company and their models.

▲Planktonne 7 hours ago | parent | next [-]

> probably not the intention, but these kinds of titles feel like they're trying to shift the credit from the person to the company and their models.

Unequivocally that is the intention.

▲ktistec 6 hours ago | parent | next [-]

To some extent they should, given he’s using models that aren’t publicly available and resources that, while not easily measured, probably run into the millions.

▲quasigod 6 hours ago | parent | prev [-]

Oh I'm certain Anthropic (and any other company) would do this intentionally, I just didn't want to accuse OP of it as I have no idea if they are in any way affiliated.

▲pradn 7 hours ago | parent | prev | next [-]

It's totally unfair. A more standard title would be:

"NYU mathematician working with Claude claims proof of the Riemann Hypothesis"

In newspapers, professors are always associated with their universities, not the commercial relationships they have.

▲ibn_khaldun 6 hours ago | parent | next [-]

> In newspapers, professors are always associated with their universities, not the commercial relationships they have.

The title is probably what it is because Levent Alpöge is not a professor at Harvard (or anywhere as far as I know) nor did he do this work on their behalf (or any academic institution's).

He's employed by Anthropic and used their technology. Obscuring that in the name of some semblance of journalistic fairness would actually be a disservice.

I think the title is vague, but associating him with a university is irrelevant to that.

▲stereoabuse 6 hours ago | parent | prev [-]

Alpöge is not an NYU mathematician, he is an Anthropic employee. You may be thinking of Buckmaster, who worked with Alpöge on Navier-Stokes.

▲pradn 6 hours ago | parent [-]

Ah you’re right, my mistake.

It would be better to lead with “mathematician”, as he’s not a random employee.

▲dhosek 6 hours ago | parent [-]

Indeed, until I saw the top comment identifying Alpoge (and after skimming over the post), I was thinking this was a mathematically-inclined Anthropic employee, not a full PhD research mathematician with a wide track record of accomplishment in the field.

▲GuB-42 3 hours ago | parent | prev | next [-]

If he is on the Anthropic payroll, on company's time, then Anthropic can claim credit for the discovery and put itself forward. In fact that is probably the only reason why they worked on the problem, for publicity.

So of course, they want the name of the company and its product to go before the person's name.

Also, I think that AI companies have the best mathematicians in general. AI is math-heavy, and they pay well, and besides education, there are probably not that many work opportunities for mathematicians, that should make AI companies particularly attractive for them.

▲quasigod an hour ago | parent | next [-]

Anthropic can claim the credit, and I'm sure they will cause that's marketing. I'm criticizing the title of this post chosen by OP, because I don't think the post should be marketing for Anthropic.

▲blackmesaind 3 hours ago | parent | prev [-]

Anthropic is a company filled to the brim with nerds obsessed with math. I'm sure the publicity isn't the only reason they worked on the problem. The public already understands that AI is good at math - to them, there's hardly a difference in significance between Navier-Stokes and GRH.

▲benmusch 3 hours ago | parent | prev [-]

Did you read the remarks? You can argue Levante has a monetary incentive to credit his employer, but the actual human mathematician reporting this result clearly gives most of the credit to the AI model

▲quasigod 39 minutes ago | parent [-]

He definitely gave a lot of credit to the model, but he also states that he found it with Claude, not that Claude found it. I have no idea if more of the credit belongs to Claude or to him, and I don't really want to speculate about monetary incentive.

Regardless, my point with the title was that a mathematician discovering this by working with Claude, versus a researcher with little to no relevant experience, using unlimited compute to discover it without much input, are two very different things. Without specifying the former, I think some would assume the latter.