[jrm4]

At the risk of sounding like an old head;

Seems to me then, priority one should be "free and open source all the models as hard as possible, so that EVERYONE can fine-tune."

(This being a subset of the idea of, free / open source is generally preferable for both freedom and quality)

[klabb3]

It seems to me this means whoever has hoarded and declared ownership of the most personal data will make the best products. Kinda like how some people liked their targeted ads because they’re more “relevant”, only now it’s not just ads but useful products. Another winner is of course platform owners like Apple and Microsoft who can scrape your data off their apps and products, even locally. This is a much bigger edge than being 3-6 months ahead in model quality.

I despise the centralization of this tech as well, and while it’s hopeful that smaller fine tuned models are better, they won’t win (or barely stand a chance) out of the virtue of openness and privacy alone. Best we can hope for is proliferation in the small-medium sized business service space - that OpenAI tokens are not worth the extra expense if open models are commoditized and effective. This was probably Zuck's plan all along – to prevent centralized gate keepers in tech that’s mainly benefiting his rivals. But the enemy of my enemy is my friend, so his actions may be the best he’s ever done for the public good.

[jrm4]

Your end point I think is exactly right.

I think your first one is getting downvoted hard because your first sentence is not at all how any of this works.

Sucking down personal data isn't JUST a bad idea for privacy, it's actually also bad for "making the best products," I think you're overstating the extent to which all that data that is stolen and sold to the highest bidder actually helps the company buying it?

[klabb3]

Ah thanks for pointing out. I don't care much for LLMs at all, but my point was simply that whoever has data, and especially personalized data, has an upper hand in making LLMs into better end user product, for those that like them. This may be underestimated right now when most dick measuring is comparing model-model not integration into a product.

> data that is stolen and sold to the highest bidder

Didn’t mean necessarily the data brokers (although that’s an interesting angle), but say Apple now has a bunch of info about your calendar, email, contacts, then clearly they have an upper hand in providing better products than an anonymous API call. Not all products need personalization but LLMs? I can think of tons of use cases.