Katana Zero
Celeste
Cuphead
Opus Magnum

Turns out having a value proposition beyond “we bundled a lot of software together that you can get on any distro” has allure.

Something like Microsoft Word or Paint is not generative.

It is standard for publishers to make indemnity agreements with creatives who produce for them, because like I said, it’s kinda difficult to prove plagiarism in the negative so a publisher doesn’t want to take the risk of distributing works where originality cannot be verified.

I’m not arguing that we should change any laws, just that people should not use these tools for commercial purposes if the producers of these tools will not take liability, because if they refuse to do so their tools are very risky to use.

I don’t see how my position affects the general public not using these tools, it’s purely about the relationship between creatives and publishers using AI tools and what they should expect and demand.

Those analogies don’t make any sense.

Anyway, as a publisher, if I cannot get OpenAI/ChatGPT to sign an indemnity agreement where they are at fault for plagiarism then their tool is effectively useless because it is really hard to determine something in not plagiarism. That makes ChatGPT pretty sus to use for creatives. So who is going to pay for it?

While I agree that using copyrighted material to train your model is not theft, text that model produces can very much be plagiarism and OpenAI should be on the hook when it occurs.

It’s not hypocritical to care about some parts of copyright and not others. For example most people in the foss crowd don’t really care about using copyright to monetarily leverage being the sole distributor of a work but they do care about attribution.

LLMs don’t “know” anything. The true things they say are just as much bullshit as the falsehoods.

Literally Bill Gates.

Who’s labor?

I’ve had to debunk this myth multiple times in this post already, I’m not sure who started it. Legumes do have all the aminos and in sufficient amounts.

Also on bioavailability, it is a double edged sword. For example heme uptake is greater than mineral iron, but your body has very little control with inhibiting uptake when you already have adequate levels. With the mineral form your body has various ways to promote or inhibit uptake. The same is true of your example of vitamin A. You can pretty much eat as many carrots as you want and suffer no ill effects, eating too much liver or taking too many liver oil supplements however can lead to poisoning.

Peanuts are cheap in the US. There’s always peanut butter too, which despite the processing required is probably even cheaper, maybe because of shelf stability and packing density.

Not just meat, but veggies as well, would get a huge boost from looking at protein per calorie. In terms of available calories, vegetables are often very protein rich, especially the dark leafy ones, but they are also packed with water and indigestible fiber. Nobody is going to eat 2000 calories of spinach, but when spinach is on your plate you shouldn’t look at it as hurting your protein intake. Also a chart based on calories that had lines down the middle for suggested protein intake and suggested intake for bodybuilding would reveal that most whole foods that aren’t fruit are completely adequate, the only reason you’d ever have to eat a lot of the highest protein per calorie foods is if a large portion of your diet was refined sugars or oils. Supplementing your Oreo addiction with chicken breast is not a good way to think about a balanced diet.

There are different skills than just physical execution. Pokemon isn’t easy because RNG or because it’s turn based, it’s easy because the NPC team compositions are awful, the AI sucks, and the game only has very lenient soft caps on grinding. A mod like Radical Red solves these things, and I’ve played other turn based games with plenty of RNG which require lots of skill.

I didn’t think it was a choreographed publicity stunt. I just know Altman has used AI fear in the past to keep people from asking rational questions like “What can this actually do?” He obviously stands to gain from people thinking they are on the verge of agi. And someone looking for a new job in the field also has to gain from it.

As for the software thing, if it’s done by someone it won’t be openai and megacorporations following in its footsteps. They seem insistent at throwing more data (of diminishing quality) and more compute (an impractical amount) at the same style of models hoping they’ll reach some kind of tipping point.

This fear mongering is just beneficial to Altman. If his product is powerful enough to be a threat to humanity then it is also powerful enough to be capable of many useful things, things it has not proven itself to be capable of. Ironically spreading fear about its capabilities will likely raise investment, so if you actually are afraid of openai somehow arriving at agi that is dangerous then you should really be trying to convince people of its lack of real utility.

MusicBrainz Picard

Amazing music tagger and batch renamer, for those of us who still have all our music as files.

You could probably just output OBS to a virtual webcam and just do a regular video call over Discord.

If you create data with the intention for that data to be used with a program then that data becomes a part of that program and you are therefore programming when creating that data.

If the protesters in these other situations were blocking fleets of ambulances and firetrucks you might have an equivalency.

Demonstrably effective

Where’s the logic in looking at something successful and picking a singular thing to be responsible? What seems more likely is you are looking for an idea you are attached to that exists adjacent to something successful. It’s like a Mormon looking for successful Mormon CEOs to then claim the company’s success is due to the Mormon work ethic. It’s like how in Whiplash the Charlie Parker story is venerated and seen as explanatory by the characters.