@HistoPol It gets even better. All Apple, Google, and Amazon announced internal projects to counter OpenAI only this past week.

What if other universities and medium-sized upstarts create a hundred other Alpacas?

Will MAGAFA have to withdraw their precious APIs, so that other models don't learn from theirs?

I don't share the dread other people feel over these developments, the very big corporations are accelerating exponentially on a collision course with their own demise.

@olavf

LLaMas ---> Alpaca

Get it? 😂

@olavf @vruz I'll happily bootstrap an AI by pirating someone else's ethically shady shit. Seems more ethical that way.

@SocialistStan

Shocking, I know! 😂

@thoughtshopper Interesting. Will read later. Thanks for sharing.

I call it all plagiarAI.

@vruz

@opendna For people who are proponents of capitalism as it was intended, that's a best case scenario.

Of course, these days in centrally planned capitalism, that person is in the vast minority.

@mmby Fast enough that they can erect new barriers to entry. But fast enough, in this case, would have to be a matter of hours, which just can't happen in the legal world.

If that doesn't happen with the speed required, thousands of medium-sized upstarts have enough time in the coming days and weeks to attempt to reproduce what Stanford did.

@pzingg It's a tall order call for anyone to punish Stanford, who are not reproducing literally any protected intellectual property works, nor their technical methods.

I think the real legal problem is for MAGAFA to erect new legal barriers to innovation fast enough that they can even begin to make a profit from this type of technology.

@Merovius @dr2chase

Alpaca also borrows quite a lot from Meta's open sourced training materials, so, yes... in a way it's indeed coming back at them.

Not that this is qualitatively an unexpected aspect of Free and Open Source software, what's unexpected and disruptive is the dramatic, possibly unprecedented magnitude of the drop of the cost of producing a comparable product from openly available materials. Near zero, as we discuss it, which wasn't supposed to happen for another few years.

@Merovius I find it amusing that none of the OpenAI people seem to have read dürrenmats the physicists. If so they would not do a surprised Pikachu now

@corbin The real insight is that I managed to condense a highly technical topic in layman's terms that anyone could understand in just a couple of paragraphs so that everyone could reach their own conclusions, even you.

I agree I did a pretty darn good job, indeed, thank you.

@pilhofer You may be able to read the paper that's referred to from the linked sources. The names of the authors are written in latin characters in the paper.

@pilhofer anything happening at any University is , from my experience.

@yetiinabox @artificialphilosopher

We're right there, roaming about he valley at the foot of the Mountains of Madness. 🙂