Prepare for That Stupid World
Key topics
The debate rages on around an experiment where Anthropic integrated a chatbot into a snack vending machine, sparking concerns about the potential consequences of such AI applications. Some commenters, like valleyer and ipdashc, question the practicality of the idea, with valleyer bluntly stating it's "stupid" and ipdashc suggesting the author is "catastrophizing" the situation. However, others, such as rdiddly and chuckadams, point out that the experiment was likely a publicity stunt, meant to draw attention rather than being a serious application. As the discussion unfolds, it becomes clear that the experiment was actually testing the AI's ability to manage the vending machine, not just interact with it, with some commenters, like sschnei8 and xp84, providing context and nuance to the original experiment's goals.
Snapshot generated from the HN discussion
Discussion Activity
Very active discussionFirst comment
14m
Peak period
98
Day 1
Avg / period
33.3
Based on 100 loaded comments
Key moments
- 01Story posted
Dec 19, 2025 at 12:01 PM EST
21 days ago
Step 01 - 02First comment
Dec 19, 2025 at 12:15 PM EST
14m after posting
Step 02 - 03Peak activity
98 comments in Day 1
Hottest window of the conversation
Step 03 - 04Latest activity
Dec 28, 2025 at 8:25 PM EST
12 days ago
Step 04
Generating AI Summary...
Analyzing up to 500 comments to identify key contributors and discussion patterns
Want the full context?
Jump to the original sources
Read the primary article or dive into the live Hacker News thread when you're ready.
I fear the author has missed the point of the "Project Vend" experiments, the original write-ups of which are available here (and are, IMO, pretty level-headed about the whole thing):
https://www.anthropic.com/research/project-vend-1 https://www.anthropic.com/research/project-vend-2
The former contains a section titled "Why did you have an LLM run a small business?" that attempts to explain the motivation behind the experiment.
Sure, but like the other guy said, that's the point of publicity stunts. It doesn't even have to be specific to a company/ad, any silly thing like this is going to sound crazy if you take it seriously and "extend its logic as far as it will go". Like seeing the Sony bouncy balls rolling down the street ad and going "holy shit, these TV companies are going to ruin the world by dropping bouncy balls on all of us". It's a valid thought experiment, but kind of a strange thing to focus on so sternly when it's clearly not taking itself seriously, especially compared to all the real-world concerning uses of AI.
I feel like he's catastrophizing the ordinary amount for an anti-AI screed. Probably well below what the market expects at this point. At this point you basically have to sound like Ed Zitron or David Gerard to stand out from the crowd.
AI is boiling the oceans, and you're worried about a vending machine?
https://andonlabs.com/evals/vending-bench
Now the shoe is on the other foot. Prepare for what happens next. FAFO.
It has always exited, but its overt forms are very much in vogue today and even celebrated publicly.
You are welcome to continue posting nonsense but the world will move forward with AI with or without you.
If only I could get any journalists or companies to actually listen to me.
[1] https://youtu.be/hUyj3d-BSh8
[2] https://youtu.be/POl7UYwBpWw
I got an S25 recently and when I search for "wife" it tries to find pictures with my wife in them. But before it does that it has to ask me who my wife is. There's no way to get it to search for the word "wife." (If I'm wrong, please tell me how.) Other text searches simply don't work either.
Sometimes it's the small ways in which the world is getting dumber.
insert obligatory throwback quote from some antique dude complaining about the youth
This has been a trope since literally the beginning of civilisation. I don’t think it’s any more true or insightful in the modern era
hmm, based on what evidence?
Or, if you prefer, based on what appeal to authority? Did you actually quote that authority properly or did you just wing it? Can you properly quote many authorities?
If you don't have good answers to those, then perhaps you have just proved the your opponents point?
Maybe there is a reason people need more compute in their key fob than what our parents/grandparents needed to pilot their ship to the moon?
Horace, Book III of Odes, circa 20 BCE
“Youth were never more sawcie, yea never more savagely saucie . . . the ancient are scorned, the honourable are contemned, the magistrate is not dreaded.”
The Wise-Man’s Forecast against the Evill Time, Thomas Barnes 1624
Some more here https://historyhustle.com/2500-years-of-people-complaining-a...
Either things have gotten continually worse for the last 3000 years or it’s just a tired trope from old men.
But if you want evidence that we're improving, I'd point out that 20 years ago, the mainstream US position was that gay people were evil, 60 years ago they thought black people shouldn't be allowed to vote, and 100 years ago they thought women were also inferior and shouldn't be allowed to vote.
We can keep going back to when people thought "slavery" and "the divine right of kings" were solid ideas.
So... if people were so much smarter in the past, why did they believe all these obviously-dumb ideas?
In any case I was looking at a longer view - maybe we have been getting more stupid in the last decade or so but who can say for sure?
1) because dude, it’s the Wall Street Journal; the entire episode should be viewed as Anthropic preparing to Ollie into an IPO next year.
2) I’m starting to interpret a lot of blog posts like these as rage bait
But I do get the point that the author is trying to make.
I just wish that there were some perspectives on the subject as a whole (AI’s sloptrod into every crevice of human life; modern technology and society and general) that don’t terminate on ironic despair.
This feels forced, there are obvious and good reasons for running that experiment. Namely, learning how it fails and to generate some potentially viral content for investor relationship. The second one seems like an extremely good business move. It is also a great business move from WSJ, get access to some of that investor money in an obviously sponsored content bit that could go viral.
Having said that, I do feels the overall premise of the blog - the world dynamics seems exceedingly irrational in recent times. The concerning fact is that irattionality seems to be accelerating, or perhaps it is keeping pace with the scale of civilization... hard to tell.
That's... exactly what the author said in the post. But with the argument that those are cynical and terrible reasons. I think it's pretty clear the "you" in "why would you want an AI" vending machine is supposed to be "an actual user of a vending machine."
The closest that I think he even gets to one is:
> At first glance, it is funny and it looks like journalists doing their job criticising the AI industry.
Which arguably assumes that journalists ought to be critical of AI in the same way as him...
Right, and neither did the GP. They both offered the exact same two reasons, the GP just apparently doesn't find them as repugnant as the author
The two reasons I believe you may be referring to from above are:
1) "learning how it fails" 2) "to generate some potentially viral content for investor relationship."
The whole of Ploum’s argument may be summarized in his own words as:
> But what appears to be journalism is, in fact, pure advertising. [...] What this video is really doing is normalising the fact that “even if it is completely stupid, AI will be everywhere, get used to it!” [...] So the whole thing is advertising a world where chatbots will be everywhere and where world-class workers will do long queue just to get a free soda. And the best advice about it is that you should probably prepare for that world.
I hate to be pedantic...but my growing disdain for modern blog posts compels me to do so in defense of literacy and clear arguments.
Whether the GP and the author offer the “exact same two reasons” is a matter of interpretation that becomes the duty of readers like us to figure out.
If we take Ploum’s words at their face...the most he does is presuppose (and I hope I’m using that word correctly) that the reader is already keen on the two reasons that `TrainedMonkey makes explicit and like the author, finds them to be stupid. While he does say that the video is not journalism and that it is advertising and that the video does show how the AI failed at the task it was assigned he does not give any credence as to why this is the case from a position other than his own.
Maybe I’m misunderstanding the concept of a “charitable interpretation” too. But I don’t think that there is one present in this post that we’re responding to. `TrainedMonkey’s comment leads off by telling us that this is what (I think) he’s about to offer in the remarks that follow when he says “there are obvious and good reasons for running that experiment”.
So my gripe is that you’re making it sound like there’s a clear counterargument entertained in this post when there isn’t. Because you overstated your interpretation of the GP comment in what looks like an attempt to make Ploum’s argument appear more appealing than it ought to be. Even though both `TrainedMonkey and myself have expressed agreement with the point he’s trying to make in general, perhaps we’re less inclined toward pugnaciousness without a well thought out warrant.
Good business moves can often be bad for humanity.
There will be no more vending machine manufacturers/operators once Anthropic masters the vending machine manufacturing and operating AI.
Running low on CandyBars is a variation on running low on WorkingVendingMachine.
Does this need an /s tag? I'm increasingly unsure.
Had a great business idea just now: A tool for staged interviews! The subject and the journalist submit an equal length list of questions. Each round of the auction they bid on questions they want to include or exclude. The loser gets 50% of the points spend by the winner to be used in the next round. Both the subject and the journalists can buy additional points at any time. I keep all the money.
Humans were just not needed anymore, and it terrifies.
Humans were never needed (for what?)
> What this video is really doing is normalising the fact that "even if it is completely stupid, AI will be everywhere, get used to it!"
Techies are finally starting to recognize how framing something as "it's inevitable, get used to it" is a rhetorical device used in mass communications to manufacture consent.
See:
https://news.ycombinator.com/item?id=44567857 'LLM Inevitabalism' 5 months ago
https://news.ycombinator.com/item?id=46288371 'This is not the future' 3 days ago
What happens in 4-5 years when we suddenly have no new engineers, scientists, or doctors?
Young people don't have the life experience to know how unrealistic these claims are, all they can do is act on the information as it's presented. It's irresponsible at best, and evil at worst.
But, the point of the article is not that you would implement an agent based vending machine business. Humans restock the machine because its a red-team exercise. As a red-team exercise it looks very effective.
> Why do you ever want to add a chatbot to a snack vending machine? The video states it clearly: the vending machine must be stocked by humans. Customers must order and take their snack by themselves. The AI has no value at all.
Like this is like watching the simpsons and being like "why are the people in the simpsons yellow? people in real life aren't yellow!!"
The point isn't to run a profitable vending machine, or even validate that an AI business agent could become profitable. The point is to conduct an experiment and gather useful information about how people can pwn LLMs.
At some level the red team guy at Anthropic understands that it is impossible by definition for models to be secure, so long as they accept inputs from the real world. Putting instructions into an LLM to tell it what to do is the equivalent of exposing an `eval()` to a web form: even if you have heuristics to check for bad input, you will eventually be pwned. I think this is actually totally intractable without putting constraints on the model from outside. You'll always need a human in the loop to pull the plug on the vending machine when it starts ordering playstations. The question is how do you improve that capability, and that is the anthropic red-team guy's job.
Having an AI run an organization autonomously is exactly the point of Andon Labs [0], who provided the system that WSJ tested.
[0] https://andonlabs.com/
https://youtu.be/5KTHvKCrQ00
It's a bit sparse on details, but it did have what in a human we would call a psychotic break.
I find this very amusing in light of OpenAI's announcement that GPT now solves >70% of their knowledge work benchmark (GDPVal). (Per ArtificialAnalysis, Opus is roughly on par.)
The economy is about to get... Interesting ;)
Is it some Viktor Frankl level acceptance or should I buy a copy of the Art of Electronics or what?
Advice welcome.
>The first thing that blew my mind was how stupid the whole idea is
Billions are being poured into LLMs. How is it stupid to experiment with them and see how they fail as opposed to ignoring that?
They weren't caught out by it, they didn't present a working solution, it was just a fun bit of research.
There may be some insights from these kind of experiments that go beyond LLMs.
I read a random comment a few days ago from someone who was saying it'll become possible for people, politicians, companies, etc to run speeches, policies, ideas, etc across thousands of LLM "personalities" to fine tune messaging and it sure seems prescient.
99.9% of social media comments fail to do this.
If the journalist was not asking the right questions, or was too obvious the article was PR it’s another thing (I haven’t read WSJ’s piece, only the original post by Anthropic)
Yes, but as stated by the Anthropic guy, a LLM/AI running a business is not. Or would you just let it run wild in the real world?
And I agree that there is a PR angle here, for Anthropic could have tested it in a more isolated environment, but it is a unique experiment with current advancements in technology, so why wouldn't that be newsworthy? I found it insightful, fun and goofy. I think it is great journalism, because too often journalism is serious, sad and depressing.
> None of the world class journalists seemed to care. They are probably too badly paid for that.
The journalists were clearly taking the piss.They concluded experiment was a disaster. How negative does the author want them to be about a silly experiment?
This was just a little bit of fun and I quite enjoyed the video. The author is missing the point.
When electricity was "the hot new thing" people tried all kinds of bizarre and seemingly insane things with it, at least from our modern perspective.
Things like:
The Electric Table Cloth, one of the first "wireless" home gadgets https://www.reddit.com/r/CrappyDesign/comments/hgemyu/an_ele...
The Schnee Bath, because what's better than a bath for healing and relaxation? An electrified bath! https://commons.wikimedia.org/wiki/File:Dr._Schnee%27s_four-...
The Solar Bath Apparatus, they say sunlight is good for your health so let's turn it up to 11 https://www.reddit.com/r/OldSchoolRidiculous/comments/15h7ce...
and there were many others whose initial designs were incredibly inefficient or dangerous but which turned into real products that we still use today after refinement. Toasters used to just be the bare heating element and left you in control of handling the toast with a long fork or tongs. The original electric blanket was a metal pan with a lightbulb in it that you used to warm up your bed. Some of the first electric razors just added a vibrating motor to a normal safety razor. The great-grandfather of electric lighters was a fireplace poker that heated up like a soldering iron so you could jam it into your fireplace to light it (and you can actually still buy these! https://www.dopogrf.com/product-p-769347.html).
So let's just laugh at the funny and quirky uses people are coming up with for AI rather than being a grouch. It's entertaining and we will probably get some really useful inventions out of it in the long run.
Since the T&C update came - of course - from no-reply@bunq.com I went to their website and quickly found out, unless I install their App again, there is no way to do anything. After installing the App, they wanted me to record a selfie, because I was using the app from a new device. I figured that is a lot of work and mostly somewhat unreasonable to record a new selfie just to have my data deleted - so I found their support@bunq.com address.
And, of course, you guessed it, it is 100% a pure AI agent at borderline retard level. Even though it is email, you get AI answers back. My initial inquiry that I decline the T&C and want to terminate my account and my data deleted via GDPR request was answered with a completely hallucinated link: bunq.com/dataprotection which resulted in immediate 404. I replied to that email that it is a 404, and the answer was pretty generic and that - as well as all responses seem to be answered in 5 minutes - made my suspect it is AI. I asked it what 5 plus five 5 is, and yes, I got a swift response with the correct answers. My question which AI version and LLM was cleverly rejected. Needless to say, it was completely impossible to get anything done with that agent. Because I CC'ed their privacy officer (privacy@bunq.com) I did get a response a day later asking me basically for everything again that I had answered to the AI agent.
Now, I never had any money in that account so I don't care much. But I can hardly see trusting a single buck to a bank that would offer that experience.
Why not? Read the article and ponder the seven to eleven distinct ways the AI goes wrong. Every AI enabled workflow will fail the same way. It even raises the question “is this technology even deserving of the title AI at all?” (If you ask me, probably not)
It lacks common sense, it lacks a theory of the world it lacks the ability to identify errors and self-correct. It lacks the ability to introspect. Now consider the thousands of people out there pushing the idea of fully autonomous AI “agents” doing work in the real world. Are we really for that? Again the exercise answers the question definitely. “No, the technology is not ready and cannot be trusted, and maybe in the current incarnation (Namely LLM based) cannot ever be trusted”
Now consider the tens of millions of people who think the technology IS ready for that. Anthropic is in the business of studying AI safety, publishing research, examining what AI is and how far we can trust it. They did that job smashingly.
> "Logan Graham, head of Anthropic’s Frontier Red Team, told me the company chose a vending machine because it’s the simplest real-world version of a business. “What’s more straightforward than a box where things go in, things go out and you pay for them?” he said."
This was a project of Anthropic's Red Team, not a product development team. Deploying the AI in a vending machine context was chosen as a minimal "toy model" in which to simulate a larger, more complex real business.
> "That was the point, Anthropic says. The Project Vend experiment was designed by the company’s stress testers (aka “red team”) to see what happens when an AI agent is given autonomy, money—and human colleagues."
Anthropic had already done this experiment internally and it succeeded - by failing to operate even the simplest business but doing so in way that informs future research about failure modes. Offering to allow the WSJ to repeat the experiment was an obvious PR move to promote Anthropic's AI safety efforts by highlighting the kinds of things their Red Team does to expose failure modes. Anthropic knew it would fail abjectly. The whole concept of an AI vending machine with the latitude to set prices, manage inventory and select new products was intended to be ludicrous from the start.