Post

I like compromises. But one place I don't believe in meeting people halfway, or 5% of the way, is the notion that AI would be better if not based on texts/images that governments have made illegal to copy. "OK, technically legal since you're not copying, but still, you should feel bad." Nope! +

It's okay to learn from everything, and train on everything. That's everyone's birthright, and we shouldn't apologize for it. Prohibitions on copying were a late invention, and a narrowly tailored one. They do not express a general ethical law that all information ought to be priced and paid for.

To avoid subtweeting; I'm responding to this theophite/scalzi convo and suggesting that -- while I agree with Mr Arson -- it might be strategic to take a even harder line by rejecting the tacit premise.

This post is unavailable.

While I agree that information should be free and everyone should be building on the shoulders of the giants that came before us, not acknowledging the people who you learned from, flagrantly ignoring their voices, and building a toll bridge to knowledge is fucking rude as hell.

Fortunately, that bridge itself is easy as heck to copy and make free. At this point pretty much everything that was gained from the initial pre-training is in fact, universally shared. Differences between models that people pay for are now all post training.

1 repost44 likes

Right now the entry fee build your own bridge is thousands of dollars, it may come down in price, but not in the near future. On a long enough timeline information trends to become cheaper.

All the raw materials to build a bridge are freely available to you and you're complaining that no one will also build the bridge for you for free?

There are a number of competitive inference utility providers. Unless you're a dedicated hobbyist or have research/business funding behind you I can't see why you would need to build your own