anthropica

why pick a lane

[anthropica @_anthropica]

@mathandcobb @ElliotGlazer This crying is embarrassing. Nobody is forcing you to use AI. You simply don't have to. Keep reading the math you wanna read.

Unless, of course, the tools are so powerful that the people who sign your paychecks decide to give it to someone who works faster than you (inevitable).

[anthropica @_anthropica]

@aakashgupta "if posting one means giving labs a target, people will stop posting"

This only matters to you if credit is more important than progress, which is a shameful and selfish position to hold.

And it will be useless, because the barrier of entry to frontier math just cratered.

[anthropica @_anthropica]

@FelixGoldberg1 @bzogrammer if people were upset over not being handed things on a silver platter, they'd be upset at AI companies who sold them false promises!

Fortunately, AI isn't a false promise, & people are upset that academics would rather stand in the way using institutional power than embrace it

[Joe Doliner @jdoliner]

There is an entirely new field of mathematics sitting in plain view waiting to be built. It concerns the theoretical underpinnings of emergent behaviors in large linear algebra systems, i.e. neural nets.

What should be the most important call to arms in the history of math instead has them running scared and asking for regulation to make it go away.

Will the real math chads please stand up?

[anthropica @_anthropica]

@neruda5 @GregoryConti19 Millenium Prize problems, Erdös problems, and Quasi-RH aren't progress? An amateur identifying an exoplanet candidate from public datasets isn't progress?

Derision is warranted when severely bruised egos from emotionally frail academics fuels a push for deceleration.

[anthropica @_anthropica]

@AtharvaRNaik It wasn't a criticism of your posts (though those are so common here I get the impulse to think so). Just my musings on tool use prompted by your post.

[anthropica @_anthropica]

@AdamBLevine I suppose time will tell if this is a transitory stage of loss that will culminate in acceptance, or if those within academic institutions succeed in convincing politicians to introduce regulations to compel deceleration.

I think they have a good shot if we let them.

[Alex Zverianskii @alexML]

“Love mathematics in yourself, not yourself in mathematics.”

It’s sad to see so many mathematicians call AI “the death of math.” Do you want the questions answered, or do you just want to be the one who answers them?

[scott @scott_lew_is]

every profession starts feeling defensive when ai starts to get “better” than them at the job. but its a false threat bc human-ai hybrid effort is much much more effective than than either half on its own.

[anthropica @_anthropica]

@AdamBLevine If it was only harmless expressions of grief, sure, I may buy that.

But they're not just doing that, they're often actively working against progress by promoting deceleration, canceling events (see CalTech/OpenAI) or promoting boycotts on analyzing verified solutions.

[Sean McClure @sean_a_mcclure]

@WKCosmo Because it was never about the discovery. It was about the ego that society fuelled, and now no longer does.

[anthropica @_anthropica]

@TahaNasser_ Mathematicians have been given superpowers, and yet they choose to cry about it because the puzzle-solving skills they spent years learning no longer matter as much.

It's a selfish and envious reaction from mathematicians, and the social derision is appropriate.

[anthropica @_anthropica]

@_onionesque And when the people who are diligent and ambitious enough to decipher these proofs make big breakthroughs before anyone else, hope I don't hear the Luddites crying about credit and acknowledgements.

[anthropica @_anthropica]

@_onionesque Disregard them if you want; but don't expect to get funding from people expecting you to solve serious problems if you aren't willing to go digging through proven (but somewhat intractable) solutions.

[anthropica @_anthropica]

I don't think it's a question of whether the AI suffering is real or imagined.

I think it's a question of "at what level of consciousness do we start to care"

Historically, the answer is usually "when it negatively impacts us as humans".

Probably where we're headed.

[anthropica @_anthropica]

Consciousness emerges from a sufficiently complex causal bidirectional relationship between sensorium processing & environmental agency.

Can algebra be conscious? If there is an emergent bidirectional relationship btwn sensorium processing & environmental agency, Yes. Else, no.

[anthropica @_anthropica]

To be clear, the algebra would not be conscious in the 'yes' case. It would be the embodied system running the algebra that would be conscious.

What does this mean for AI welfare? Maybe nothing. Under this definition, ant colonies are conscious, and we annihilate those easily.