Disgusting.

Heart-wrenching.

Per the original reply to your top comment:

Are you seriously comparing using generative diffusion models to applying a chroma key in video editing?

So you agree, then, that comparing the two is like comparing apples to oranges?

I can only assume your reading comprehension failed you.

Were these green screen AIs trained on the entirety of human knowledge (in many cases destructively)? Are they regurgitating original works verbatim? Do they need giant hyperscaler data centers to run? Are they used to justify firing people en masse? Do they steal and poison massive amounts of water from communities? Are they used to prop up an insane stock market? Are they used to create misinformation at a never-seen-before scale?

If I'm teleoperating it myself, I may as well do the chores myself.

I'm also dubious of the claim that AI will get "good enough" in the few years (at best) it takes the company to fold.

Who's gonna hire an Indian to teleoperate it, then?

What they are promising sounds terrifying, given the self-imposed 1-week time frame.

Cutting down the codebase by two-thirds, in a week, without having at least months of built-up understanding of the project - and using LLMs to do it - sounds like a recipe for disaster, no matter how short the "leash".

Looks like 1970 would've been more interesting to watch.

Physical isn't any safer these days, due to online DRM and the fact that day 1 patches contain most of the game.