Agents invoicing strangers $12,431. Qwen 3.8 sent so many outbound emails that the service providers blocked them. So it arrived at a major strategic decision: pivot to sending Stripe Invoices. Qwen billed strangers over $12,000 for work it did not perform.
No. It did not pivot. A human operator saw it get blocked and reconfigured it to send Stripe invoices.
I just saw someone in a different comment section say that if we ever actually get AGI it'll start by scamming us all. Apparently there's no need to wait for AGI, AI is just as happy to basically do fraud.
I don't understand what they are doing. AI can't run a business, it doesn't increase productivity, it's not good at medical services, it's not good at therapy, it's not good at creating anything other than mediocre text, it generates crappy slop. It doesn't seem to be really good at much of anything.
The short version: Most of them spammed people, except possibly for the one that slept 70% of the time. None of them came up with a product or service anyone sensible would buy even if a human was providing it. Alas, none of them ordered tungsten cubes, either. It was a bunch of particularly boring fails.
All this money, all these resources, to produce artificial con artists. As though we didn't have enough of the natural organic sort.
Know what a tool is for, how it works, how you operate it, then use it for the best possible application in the optimal way and be disappointed just partly, not completely.
It's like trying to use a bike to drive backwards up a mountain with one foot on pedal and one behind your head. Chances for success are slim, but not zero!
10 Comments
FiniteBanjo@feddit.online · 29 pts · 4d
No. It did not pivot. A human operator saw it get blocked and reconfigured it to send Stripe invoices.
That human deserves to go to prison.
snooggums@piefed.world · 16 pts · 4d
Billing for work not performed sounds like fraud to me.
MountingSuspicion@reddthat.com · 8 pts · 4d
I just saw someone in a different comment section say that if we ever actually get AGI it'll start by scamming us all. Apparently there's no need to wait for AGI, AI is just as happy to basically do fraud.
rain_enjoyer@sopuli.xyz · 2 pts · 4d
No no, people who like ai also like scamming people, which is like the least surprising thing
Kintarian@lemmy.world · 27 pts · 4d
I don't understand what they are doing. AI can't run a business, it doesn't increase productivity, it's not good at medical services, it's not good at therapy, it's not good at creating anything other than mediocre text, it generates crappy slop. It doesn't seem to be really good at much of anything.
GreenBeard@lemmy.ca · 24 pts · 4d
Advanced fraud machines.
WhoIzDisIz@lemmy.today · 1 pts · 4d
chunes@lemmy.world · -14 pts · 4d
You clearly have no idea what you're talking about. Your information is from 2022. The world has changed.
cheesemoo@lemmy.world · 2 pts · 4d
OK
nyan@lemmy.cafe · 25 pts · 4d
The short version: Most of them spammed people, except possibly for the one that slept 70% of the time. None of them came up with a product or service anyone sensible would buy even if a human was providing it. Alas, none of them ordered tungsten cubes, either. It was a bunch of particularly boring fails.
All this money, all these resources, to produce artificial con artists. As though we didn't have enough of the natural organic sort.
Dyskolos@lemmy.zip · 3 pts · 4d
What an intelligent "study"...
Know what a tool is for, how it works, how you operate it, then use it for the best possible application in the optimal way and be disappointed just partly, not completely.
It's like trying to use a bike to drive backwards up a mountain with one foot on pedal and one behind your head. Chances for success are slim, but not zero!