Hacker News new | ask | show | jobs
by yokisan 36 days ago
One would think Anthropic could point Mythos at this to solve the reseller problem outright:

- Purchase multiple accounts via resellers

- Send messages that contain a UID

- Capture these in Anthropic's logs

- Shut down account. Use any metadata to identify related accounts

/loop

5 comments

Maybe Fable is not as capable as thought?

On the one hand they talk it up as world ending and on the other hand they can't manage bot accounts on their own service.

I want to hear how this can be rationalised.

From the article "every layer of control frontier US AI companies have added (geoblocking, phone verification, credit card requirements, and now live biometric KYC checks) has produced a corresponding layer of evasion infrastructure".

No system is foolproof. They'd have to be willing to throw out some % of good customers along with the bots. Amazon can do that because they have a monopoly already. Anthropic can't risk it when they're trying to grab market share.
In this case, being distilled is sort of existential to them. The false positives would just be losing some revenue (depending if profitable, not even losing profit).
One would believe a model scoring this high on SWEBench could maximize F1 score for a precision recall problem easily. What's the missing part?
> One would think Anthropic could point Mythos at this to solve the reseller problem outright

You're assuming Anthropic want to stop it.

I think it serves their interests more to be able to release stories like this from time to time, to feed to the US government, in an attempt to get the Chinese competition shut down.

No, it more likely serves their IPO story with growth metrics.

One doesn't ban a large chunk of the user base before IPO, because "Well, if you dig into the numbers and discount the drop from banned accounts..." is less impressive than rocketship growth (even if fraudulently fueled).

Expect more effective measures to be deployed post-IPO.

This only shuts down the account you have bought in the first, plus a few others if it is shared.

> Use any metadata to identify related accounts

How does that work? I think this is the most important part to have an impact on the „thousand“ bot accounts.

Rather than go after accounts, you’d go after the hosting facilities. The hosting companies are more likely to be able to identify related accounts so working with them will give more scalable results.
They probably will route you through different accounts. So with a single account you should be able to hit hundreds-thousands of accounts.
Especially if you did this in a sting-like manner and used subtle compound UIDs (read: unique combinations of innocuous words) then banned every account found at a later time.
They could be doing this internally and want to see if they can downright eliminate these loopholes before bringing Fable back.

I don't care how they do it, I just want to use Fable again.

This, just like blanking out a football stream for a split second to binary search and find IPTV rebroadcasters, is far too good a solution. Suits prefer to make it seem like their job of fighting "misuse" is hard, justify their budget, continued existence of the trust & safety department, face scans, etc.