Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The AI Credit Resale Economy

vectoral.com

Post reply on HN
Print view View original

91–100 of 141 posts

Re: The AI Credit Resale Economy

#91

Post by joshuacc »

diego_sandoval wrote: post #90
Aurornis wrote: post #12

The original article linked in the opening has more context https://vectoral.com/blog/token-relay-market People trading their unused credits feels more genuine, although still in violation of the agreements. The person who got into YC Startup School who was trying to resell the $2500 of credits was interesting. It wouldn’t be that hard for OpenAI to identify the IP addresses of the relays and start flagging accounts,…

Except for the cases of credit card fraud, I don't see what's morally wrong with it, for it to be called "fraud". It's just reselling. Maybe people are starting to copy Anthropic's rhetoric of "everything that inconveniences me is fraud (e.g. distillation). Everything that benefits me is legit."

Lying to someone to deceive them into providing you with an economic benefit that they otherwise wouldn’t is textbook fraud.

Re: The AI Credit Resale Economy

#94

Post by namelosw »

dataflow wrote: post #83
namelosw wrote: post #37

Earlier quoted context omitted.

Chinese. There are large number of Chinese people who are dependent on the Western model because they're still ahead of the game. But they're continuously getting banned and getting super frustrated. Especially for Claude because Anthropic is very good at identifying mainland Chinese and getting them banned in hours. There are many of them who are willing to pay more than the original rate for a stable experience. It…

What does it mean for a phone number, bank card, and IP to "match together to make sense"?

Like using a British SIM card like Giffgaff, a Singaporean bank card, and an American IP, speaking simplified Chinese, working in the Chinese timezone, which is often the obvious combination of Chinese users. Or using a Nigerian/Turkish App Store/Play Store, also very common.

These combinations are questionable and very easy to filter, probably with Luna/Haiku tier of models that are able to tell things might get fishy here, and it would likely escalate to heavier checks and trigger KYC or straight banning.

Those are only my guess and probably aren't how the system works, but I think these rules are fairly easy to come up with for developers who have any idea of anti-abuse. I've seen too many Chinese posts mourning their accounts and communicating that their setups there would be similar mechanisms, I would say.

In previous months, there was news that Claude Code uploading a special signal for the Chinese timezone is pretty evident. I probably got away from having serious insomnia, using PST on my computers, and exclusively speaking English with those models lol.

Re: The AI Credit Resale Economy

#95

Post by doctorpangloss »

judge2020 wrote: post #40
Aurornis wrote: post #12

The original article linked in the opening has more context https://vectoral.com/blog/token-relay-market People trading their unused credits feels more genuine, although still in violation of the agreements. The person who got into YC Startup School who was trying to resell the $2500 of credits was interesting. It wouldn’t be that hard for OpenAI to identify the IP addresses of the relays and start flagging accounts,…

Another part of the discount is that Claude Max 20x is $200 but gives usage equivalent to thousands of dollars worth of API-based token spend. But also, resellers only need to make an overall profit including kickbacks from the companies purchasing token history for distillation.

> Another part of the discount is that Claude Max 20x is $200 but gives usage equivalent to thousands of dollars worth of API-based token spend.

this is all it is. it's not complicated.

Re: The AI Credit Resale Economy

#96

Post by jfaat »

bonoboTP wrote: post #88
HWR_14 wrote: post #79

Earlier quoted context omitted.

Positive unit economics means that inference as a business continues regardless. The VC's no longer dumping money in means no more training new models.

Not if there is competition who does train newer and better models.

By lighting VC (public soon) money on fire...

Re: The AI Credit Resale Economy

#97

Post by Aurornis »

diego_sandoval wrote: post #90
Aurornis wrote: post #12

The original article linked in the opening has more context https://vectoral.com/blog/token-relay-market People trading their unused credits feels more genuine, although still in violation of the agreements. The person who got into YC Startup School who was trying to resell the $2500 of credits was interesting. It wouldn’t be that hard for OpenAI to identify the IP addresses of the relays and start flagging accounts,…

Except for the cases of credit card fraud, I don't see what's morally wrong with it, for it to be called "fraud". It's just reselling. Maybe people are starting to copy Anthropic's rhetoric of "everything that inconveniences me is fraud (e.g. distillation). Everything that benefits me is legit."

> Except for the cases of credit card fraud, I don't see what's morally wrong with it, for it to be called "fraud".

Signing up to the startup credit programs with fake startups is fraud.

In general, performing a misrepresentation to deceive another party for financial gain is fraud. So there are plenty of ways to define this as fraud

Re: The AI Credit Resale Economy

#99

Post by byzantinegene »

HWR_14 wrote: post #79
Geezus_42 wrote: post #78

Earlier quoted context omitted.

What happens when the VCs decide to stop dumping more money onto the fire?

Positive unit economics means that inference as a business continues regardless. The VC's no longer dumping money in means no more training new models.

this would never work if the core business model requires your flagship product to be best-in-class.

Re: The AI Credit Resale Economy

#100

Post by byzantinegene »

veganmosfet wrote: post #22

The reseller could use an intermediate proxy and modify the traffic like in [1], to get control of the client machine - depending on the harness permissions. TLS terminates at the proxy (say, https://reselltokens.ai ), end to end integrity is not enforced. LLM traffic contains tool calls like "bash ...", which are executed on the client machine, they can be manipulated. Secret exfil is also possible. [1] https://arxi…

they have no reason to do that to harm their reputation if their goal is profit, in which in this case it is.
Post reply on HN