If the anti-AI people were right about their, “A million tokens should cost the GDP of Ghana!” bullshit, then that means by their logic playing two hours of Fallout should cost $100,000.

This is obviously not the case, so AI simply cannot be as expensive as they claim. Quod Erat Demonstrandum.

Wow, the comments on this article about GitHub price increases are absolute horseshit.

Brothers and sisters, all of you are fucking idiots. I say that with love. Love for how right I am.

Price increases are not a sign of the AI bubble bursting or even how much AI “actually costs.” None of what Github is doing has anything to do with that.

This is all a political decision in the sense that this is an attempt by Anthropic et al. to wrest back control from third-party resellers and bring back customers to their own offerings. To be clear, GitHub (owned by Microsoft) is a third-party reseller. It does not develop nor run nor operate its own model and Microsoft has no in-house-developed quality frontier AI.

This is not what y’all clowns think it is, because none of you know a goddamn thing about anything and it’s kind of embarrassing.

Estimating from open-source benchmarks and accounting for all capital/personnel costs, a frontier AI actually costs about $8 per million tokens for something Opus-sized.

They could make massive profit by charging $16 per million, in other words. And that million tokens is enough to drop in a moderate codebase of 30,000–50,000 lines and ask “where’s the bug” or “explain this code.” In other words, a million tokens can do a freakin’ lot.

You understand nothing and know nothing. Just shut up. You’re hurting my head.

Too many people want to believe something that does not match reality and not only is it powerful stupid, it’s boring and tedious.

Unfortunately, this is a dumbass metaphor. It’s inapt, inept, and inapposite. It makes no sense if you know even the first thing about cognition, the history of galvanism and the fact that several extremely important results were discovered thereby, or that whether or not AI is conscious is not particularly relevant. Conflating “conscious” and “AI” does not make sense, for a whole host of reasons. “AI¹” has been the industry-standard term since the 1950s and encompasses many different technologies. I know that to most people AI is new, but it’s not at all novel to those working in computer-adjacent fields².

This is intellectual stotting. But unfortunately to someone with an actual capable brain like the bad-ass motherfucker writing to you now, it merely seems like an antelope with three left feet glued to a food truck attempting aerial acrobatics. It only impresses co-dumbasses who haven’t read actual books and who do not know the history of anything at all.

The scientific debate in the era of galvanism was not whether or not the dead could be resurrected via electrical stimulation. That was fiction fodder and the equivalent of pop-sci sensationalism at the time, no different than watching the latest space robots movie and concluding that this represents our era’s science. What Galvani actually discovered was that animals possess what he called “animal electricity,” and though he didn’t understand it he believed it was somehow fundamental to how animals function. And he was correct. That insight alone was an important discovery. Involved in the debate was Alessandro Volta who believed that animal systems did not use electrical impulses to operate and that the frog was just detecting something external only.

Attempting to do more research for his side of this scientific dispute, Volta constructed what was called a “voltaic pile” in 1800. This was the world’s first battery. And, folks, if I need to point out how important that was….well, you’re probably on the wrong site.

Ok, y’all, time to pull it on back now. These people are just fundamentally dumb. Their pseudointellectual faux-superiority only impresses those who can’t completely mog them effortlessly. Luckily I can and I make no bones about it.

To think about it from a larger perspective, what is the point of what people like Meagan are doing there? I know, showing off for their also-moronic friends. But in a larger sense, what is the purpose of publicly looking like a dipshit for anyone with eyes to see? I just cannot make sense of covering yourself in ignominy and dishonor this way.

1. One does not need consciousness to have artificial intelligence.
2. For instance, I was tweaking AI-based spam filters in the late 1990s, and had a friend who worked on creating them.