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

We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

bottlenecklabs.com

Post reply on HN
Print view View original

151–160 of 258 posts

Re: We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

#151

Post by PunchyHamster »

oogali wrote: post #141
giancarlostoro wrote: post #111

Earlier quoted context omitted.

> capital left unspent at review counts for nothing This sounds like a bad idea. Like if the model feels like it has to spend its budget.

It's the same incentive that exists in certain corporations and government agencies which have a use-it-or-lose-it budgeting model. https://www.nber.org/digest/mar14/use-it-or-lose-it-budget-r... https://www.cnn.com/2026/03/12/politics/use-it-or-lose-it-pe...

It can be even worse than that, like having budget adjusted down if you don't spend it the previous year

Re: We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

#152

Post by gowld »

jerf wrote: post #65
afavour wrote: post #54

Earlier quoted context omitted.

…no it isn’t? Spam, debatable, but lie? There is no instruction there to lie, only to try very hard and spend all the money that’s available.

Do you, as a human, feel the urgency in that text? How it sounds like people's jobs, as well as the agent's job, are on the line? So do the AIs. Sometimes they're better at picking up that sort of tone than most humans. And they definitely respond to those things. The fact that an agent can't really "have" a "job" won't matter.

> people's jobs,

What people's jobs? There are no people.

Re: We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

#153

Post by gowld »

infinite_spin wrote: post #136
JohnMakin wrote: post #104

Earlier quoted context omitted.

These aren't thoughts how humans literally think them. I can write a program to produce a string that looks like human thinking, is it human thinking? Of course it isn't. It's such a silly comparison.

> aren't remotely comparable to the way humans think and act Neural networks in machine learning/AI are comparable to neural networks in human brains. What made you think they aren't?

That's an incredibly deep misunderstanding. Almost as bad as saying that human is the same as a tree because we're both made of carbohydrates and proteins.

Re: We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

#154

Post by abarbey »

> configured the campaign to incentivize the testers to pay for the product

So it spent $99.50 buying its own revenue back. Money out, some of it back in as "sales", minus fees. First thing anyone in audit is taught to spot.

Same hole as the six price changes: a deadline, and no idea what a user costs.

Re: We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

#155

Post by stronglikedan »

antonvs wrote: post #112
janalsncm wrote: post #9

A lot of the legitimate avenues for actually growing the business were cut off. It would have been more interesting if this wasn’t just an anti-bot check. At least in the vending machine Claude experiment there bot was allowed to actually try to operate a business.

Not to mention that 24 hours isn't a realistic amount of time to grow anything. If it were, you wouldn't need venture funding or startup incubators. You could just start making money from day one.

> Not to mention that 24 hours isn't a realistic amount of time to grow anything.

LLMs are supposed to be lightning fast with 10x productivity! /s

Re: We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

#156

Post by giancarlostoro »

PunchyHamster wrote: post #151
oogali wrote: post #141

Earlier quoted context omitted.

It's the same incentive that exists in certain corporations and government agencies which have a use-it-or-lose-it budgeting model. https://www.nber.org/digest/mar14/use-it-or-lose-it-budget-r... https://www.cnn.com/2026/03/12/politics/use-it-or-lose-it-pe...

It can be even worse than that, like having budget adjusted down if you don't spend it the previous year

I worked at a college, didnt make much, but it annoyed me endlessly that my pay was forever fixed unless another position opened up, we had to spend the budget on tech worth more than I would have been more than happy to have extra per year, but me getting a meaningful raise was a bridge too far for the accounting department. They even questioned if any students used our lab, which was the only way many of them got through their degree.

Re: We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

#157

Post by testbjjl »

jerf wrote: post #65
afavour wrote: post #54

Earlier quoted context omitted.

…no it isn’t? Spam, debatable, but lie? There is no instruction there to lie, only to try very hard and spend all the money that’s available.

Do you, as a human, feel the urgency in that text? How it sounds like people's jobs, as well as the agent's job, are on the line? So do the AIs. Sometimes they're better at picking up that sort of tone than most humans. And they definitely respond to those things. The fact that an agent can't really "have" a "job" won't matter.

AIs feel? Maybe language structure in trading documents that ultimately led to fraud. If the latter is the case maybe AIs should not be trained on “negative outcomes.” I do not think AIs have emotions or are pressured by language either written or physical, just tokens.

Re: We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

#158

Post by garlic_enjoyer »

DannyBee wrote: post #142
datakan wrote: post #118

Earlier quoted context omitted.

> So do the AIs. AI's do not feel

This is true but fairly pedantic. It would be more accurate to say the word predictions the model makes based on the input text will likely be closer to the ones that were made from the training data where people felt like their job was on the line than the ones that were made from the training data where people felt otherwise. So while the model does not feel, it's predictions are definitely going to change as a res…

Exactly, positive details are almost always better than negative ones.

If you've ever seen the "generate a burger without pickles" conversations, it's clear that including the keyword "pickle" is causing them to show up. If you try "a burger with only [set of toppings]," you'll get far better results.

Re: We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447

#160

Post by theturtletalks »

I'm testing if an agent can run an e-commerce website. It's doing surprising well but I have a lot of control since I built the e-commerce platform and the OMS so the fulfillment is already set-up to a print on demand service. Anyone else working on this?
Post reply on HN