Read up on Kathryn Tewson's investigation of them. They're 100% a scam, they were never legitimate, they weren't even trying to use actual AI originally, and nothing they say is true. They've already been fined by the FTC for some of this.

This is like Slashdot running a piece on how a nigerian oil exec has twenty million dollars to donate to you. Absolutely unconscionable.

Every basic income pilot program does this, it's not surprising anymore. The first time it increased employment it was counter-intuitive, anyway, but ten or twenty experiments in, it's just not weird or surprising anymore.

this doesn't sound right, there's a bunch of benchmarks that i see in lots of announcements/model cards/whatever. some of the oldest ones aren't in use anymore because everyone aces them and they're boring, but a lot of these are sort of standard things that are the same benchmarks openai and meta used in their recent announcements?

there's a lot of standard "evals" that things are scored on, and the claim is that it performs better on a lot of those tests of functionality. that seems like a really easy and straightforward thing to find out, since it's right there in the article and in all the other news coverage about this.

Your theory that there are "far more slackers" is interesting, but not supported by data. Your anecdote isn't really related.

You're right that in theory this could support outsourcing abroad, and having coworkers in quite a few time zones, I'm actually pretty hyped about this having good potential in the future. That said, time zones are sometimes important...

A person I know had a story once, I believe about Citi, where they had a phishing message and wanted to report it to the bank's fraud/security people. And the person they got to told them that to see whether an email is legit, look at the email and see if it has a citibank logo, because that's how you can tell it's legitimate.

If my fuzzy memory of years back is correct, and that was indeed Citi, then yeah, they should absolutely be on the hook for some amount of this fraud.

(source: https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Finfosec.exchange%2F%40Goss...)

Been watching the Post Office scandal drama that ITV made. It’s really good.

One big notable so far - they introduce the CEO in episode two.. skipping over the CEO before her, who was actually in charge when all this kicked off.

He went on to be CEO of ITV.

They censor lots of things they disapprove of. Their ToS allow them to censor on lots of topics, and they do.

If they were not censoring other things, they might have a better defense. There'd be a real argument there, but it'd at least be an actual argument about the facts. The notion that they have an "anti-censorship stance" is, however, nonsense.

The key point here isn't that he wrote a fic, and it wasn't "a loving homage", he was claiming that he owned the rights to LoTR and suing over it. His suit against Amazon was what brought him to the Tolkein estate's attention.

I think if he'd made the fan work and not actively gone trying to pick a fight with them, they'd likely not have cared.

I don't think any other part of our infrastructure gets described in terms like "lost $6.5 billion in the last year". Normally we talk about what it costs to have infrastructure, because we're paying to have a thing we need. We don't talk about the school system in general "losing money", we talk about "what we spend on education".

That said, it does make sense to think about this somewhat in the context where a for-profit business is using that infrastructure and not really paying for the costs of providing that infrastructure.

I know more than one person who was a paying customer, and then suddenly their account was locked, and they never found out why. And I don't think this is a case where "they know what they did". I think they have no idea whatsoever, and Discord refused to even hint at it. Did they break a rule? Maybe! Will they ever be told which rule? No! So there's nothing they can do to change their behavior to fix it, and no reason not to just make a new account and use that instead.

The main effect is, there's no circumstance under which I'll ever pay Discord for anything, because multiple people I know got cheated out of money they paid to the service and never even got a hint as to what they supposedly did, let alone a chance to see what the "evidence" was, or dispute a finding.

Given that it will just make stuff up, it really seems to me like this is probably criminal negligence, and I really urge you to think this through: If it confidently hallucinates the answter that x is not in y, and x is actually in y, and someone gets hurt or dies, who is going to be liable for that error? "Pretty good" is not good enough for things that can kill people.

The degree to which it can confidently hallucinate plausible-sounding but completely wrong things makes me very disinclined to use it, or anything based primarily on this approach, because that's significantly anti-useful. To use it, I'm going to just have to go do all the research again myself anyway, or risk being horribly wrong. I guess it could be entertaining to play with, but after watching a bunch of people confidently relying on things it said, I'm pretty sure it's just poison. To say nothing of the various copyright-type issues, etcetera.

I just want clearer guidance on how to change my face and fingerprints when they get compromised.

It is a shame that we only have something like 150 years of research showing that "crunch times" actually hurt net productivity, not just quality, and don't actually get things done sooner than having people get enough rest so they are performing at their best. As a result of the comparatively scarce entire body of literature in this field, all of it publically available, the information needed to prevent disasters like this has been extremely hard to get, being available only to anyone who can read, anyone who has ever worked, or anyone who has ever actually met a human being.