These are great screenshots. Makes his posts super readable.

finally found the status/2114913470017860071
he is getting grilled by his hypocrisy

• tomjank/status/2115200871701372971

there were more, but he’s deleting his responses live

I am not sure that the market will bear them passing on those costs. They are already having trouble converting free users to paid tiers, and those paid tiers are not even profitable for them at this time.

Large language models are easily reproducible, that's why they're buying up all the computer parts to makesure you can't afford run one on your own hardware. Luckily I don't use these LLMs often enough to justify even using it for free

This is why they want the “AI revolution” to make computing too expensive for anyone but them. It’s about way more than control of copyright, more than seizing the means of production for themselves: they want to be able to control narratives, censor the press (established journalism and independent blog/forum posting and socials), and monitor individuals and groups that may oppose them or present dissenting viewpoints. They want summaries they can control, they want one click censorship, and they want us to rent the privilege from them, creating a permanent information-peasant class. They want insider information on markets, they want to know voter sentiment, they want to sway investors and politics to their whims via system prompts.

We already saw Elon try it (with hillariously bad results) when Grok wouldn’t shut up about “white genocide” in South Africa. Elon is just such a spectacular failson grifter that his plan had no nuance and was plainly obvious.

But now we know for sure.

Seize the means of (token) production!

A good reason to run local.

Not letting people do things is a great way to dissuade customers from using your hyperscaler service. Except we don't want anyone using AI to make slop that denigrates others or involves information hazards.

That might be why Musk and Zuckerberg say don't in their ToS but yhen let you do it anyway; tokens are tokens are profit.

But local brew will even do those things anyway.

Why would they want to rexouoe their losses? They haven't yet.

You can be happy that corpos are getting a taste of their medecine and sad that we are burning the planet to decompile Mario and Halo and mix them together.

Well I own no Microsoft or epic products and I am happy, so it seems to be working.

Technically speaking, a cleanroom reverse-engineered program isn’t illegal (this is how IBM PC clones were allowed to exist), at least in the US.

On the other hand, since AI cannot hold a copyright, the prompters cannot sue if, say, Sony or Microsoft just take their code and sell it.

Most decompiled games need a copy of the original game for assets (graphics, textures, etc) since those wouldn’t be recreated via AI, and shipping with them would be copyright infringement.

Purely generated AI content can't be copyrighted.

However, mix some human effort in, the combined work can be copyrighted. In theory if you can prove a specific thing is purely AI output, it is fair game. Good luck doing so with proof.

The cases cited in this argument all had the would be rights holder stating flat out it was untouched by human hands. Once they did anything to it, the result would be "transformative" and the AI ness won't count.

Also in this case, the games being decompiled were pre LLM. An LLM clone doesn't spare the clone of legal issues. A company tried this argument that the knock off legal papers were AI and this fair game and they lost. When Getty sued, the counter argument was that output could be infringing, but it would be the responsibility of the operator.

Conveniently for the rich people, they have rights to make knockoff machines by ingesting everything without permission, but copyright of course still counts for their purposes.

Trademark

I've been thinking about this for updating out of date mods

I’d imagine that the context window you’d need to perform such a feat would be so large that it would be impossible, even with all the compute we are currently throwing at AI models. It would at best be financially impractical.