There’s no such thing as political neutrality, all information that a model could be trained with will bias it to favour one or another viewpoint in its output, since it will be determined from whatever the sequences of words were which were fed in. Where do you imagine you’d get politically neutral training data from?
that’s ok you can still generate a CA cert
They’re just quoting the article… Which you’d know if you bothered to read at all, or even just look at the fact that it was quoted in their comment.
they have an interest in filtering out ads they weren’t paid for tho
Yeah, hence I couched it in allegedly. Ultimately we don’t really know. But even if it were true, it wouldn’t really mean anything for their profitability as a company. They have to keep improving the models if they hope to keep what customers they have.
Nobody has extremely good data about this because it’s in their interest to make that as opaque as possible, but the reporting I’ve seen suggests that it is, if not the majority at least an extremely large fraction of their customers which are selling AI services, and even where it’s used for in-house software development that is largely bankrolled by VC hoping to gain a competitive advantage with AI. It doesn’t really appear that the API pricing offers better cost/productivity than just hiring junior devs (lol).
They are (maybe, allegedly) “profitable on inference” i.e. the API pricing is higher than the actual compute they use just for the input/output tokens for the query. That excludes, though, training, capex, staff, etc., which… is by far the vast majority of their actual expenses.
Those enterprise customers are companies like this very one, who then have to heavily discount their rates or nobody wants it, (and as we see here, even then) so that’s not a solution for anthropic.
Cringe conspiracy crap. Attacking the vibes because there’s not really anything substantial to attack. Who cares if he’s arrogant? He’s a doctor setting public policy, not my friend. If you fall for an article like this you are extremely gullible.
Since it was probably AI output it’s fairly likely people will be able to do that anyway
Did you know that sometimes girls have sex with each other too?
Wait, does that mean bigoted lesbians are dishonourable plumbers?
Yeah, I guess. It’s emotionally satisfying with the context removed. But I’m not really willing to concede that to them. Certainly not when it comes to art, and honestly not really either when it comes to closed-source “clean-room” reimplementations of GPL software, however difficult to enforce that might be. I do think that software IP laws are generally bullshit though and should in an ideal world only ever protect the public interest and public goods, and never private ones, so I’ll give you that.
You’re completely right to push back — you told me to destroy the refinery in the other country but I destroyed the one in our own country instead. That’s a serious error and I won’t do it again. If you like, I can try again with a more advanced targeting routine, just say the word.
>looks at code
> AGENTS.md, CLAUDE.md, "autonomous implementation plan"
Not cleanroom.
it’s almost like you have your answer already
Basically the idea is “if they’re coming back anyway, we might as well make it easy for them”.
What exactly is the non-predatory half of that?
Right, which in English is mostly American media and social media, which is not at all politically neutral. Even if it represented the average of all viewpoints in society the output would probably be, for example, pretty racist (some people aren’t very racist, some are, the average would be a bit). But it’s not even that, since some viewpoints are far more represented than average in the corpus. That’s why that second stage is needed, because otherwise models would output some things which would be pretty embarrassing to the companies involved. Therefore I don’t really see how it could possibly be politically neutral.