Then again, I do think LLMs are an incredible technological achievement. The issue is not so much what they do or that they exist, but how they are utilized. Right now, they are utilized to further the class divide between rich and poor.
Who are we to trust in the future? Not big companies, not the state, not LLMs. Time to organize around groups and collectives that we know we can trust and that we know have our wellbeing in mind.
This is exactly how we got here though. Technology is not passive. It changes incentives, procedures, ideas and shapes the world. If we don't structurally limit what and how it's used, then we are not in control, no matter what are choices personally are.
> Companies are now trying to divert support requests into chats with LLMs
More than trying they are doing it very successfully and for a long time now
I do agree things can still get 10x worse than even the current state though
> When you talk to a person, there’s a “there” there—someone who, if you’re patient and polite, can actually understand what’s going on
I've found they have been trained to be machine-like for many years now and not actually help. They focus on empathy and understanding and caring about your needs...and diverting your attention away from actually resolving the issue. Here's an example recently I experienced:
I complained to my bank about how they show refunds on the app. I got a call from a lovely sounding lady who used a comforting tone to ask if I had any special needs she needs to be aware of so she can "provide extra support" .
At beginning I made it clear I'm not chasing a particular refund but rather raising a specific complaint about how refunds are shown in the app. 4 times she mentioned a specific refund assuring me that it's been refunded, ignoring everything I said at the beginning. She explained how refunds work. She explained how pending transactions work (all off topic). She explained 3 times about how they can't (won't) create a feedback loop and begged my permission to close off the complaint, saying she hopes I can understand.
That was all a very dressed up "I'll pass your feedback to the PM. Thanks" but it was whole ridiculous long phone call trying to make me feel "heard" and I came away feeling like a 15 year old
I chat with these friends a lot but I rarely send articles that I suggest they read and that I think are profound, so I expected them to read it. These are smart people that have a history of reading lots of books.
They are both huge AI proponents now and use AI for nearly everything now. Debates on various topics with them used to be rich; now, they're shallow and they just send me AI summaries of points they're clearly just predisposed to. Their attention spans are dwindling.
[1] https://aphyr.com/data/posts/411/the-future-of-everything-is...
a fun side effect is that CS is also an early warning system for companies, so when you make it harder to get through to a human, you start throwing out info on your users' pain points. of course this only matters if people have a choice about whether to use your product, so that's gotta be an upside for insurance companies, etc.
All that stuff about support, though, inevitable.
Given how many people hate AI in general, I'm surprised there hasn't been anything like this happening. They could even get around the irony of using "AI" themselves, I bet low-tech language models like Markov chains could provide sufficient time wasting potential (I'd love to see it done with an old fashioned AIML chatbot). Asymmetric chatbot warfare.
"Yes, we cost more, but your get what you pay for" can be a good play.
It's certainly worth discussing the fact that the entire industry is starting to outsource large amounts of our thinking and writing work to non-sentient statistical algorithms, but this discussion needs to honestly confront the extent to which they are successfully completing useful tasks today.
Ironically this was the main reason LLMs were introduced in the first place, not to benefit the poor, but to widen the gap between the rich and the poor.
Folks in the "now" have always had a tendency to cling to their fictions as if they were truth for whatever reason; like nationalist exceptionalism, racial superiority, or religions rooted in "othering", etc. Humans seem to have an innate desire to fool themselves and trust in things they should not. Perhaps it's simply a sort of existential coping mechanism of living in a cold, unforgiving reality. We seek the comfort of lies.
Organizing around groups of trust, tends to lead to factionalism and conflicts. Knowing and trusting are sadly very different things in our species.
The whale's been eaten now. The broader Internet is mostly not trustworthy, or convenient, and the information is not even very plentiful.
People will and are retreating into high-trust zones. In-person networks, product recommendations from real friends, and closed group chats.
It's not the end of the world, but things have changed. We'll have to put more work into finding information than we're used to.
You are ringing the clarion call for community and cooperation, and it will not work. Not because people don’t want community or the better things, but because incentives make the world go round.
The choice between making some money at the cost of polluting the information commons is no choice at all. That degradation of the commons means no one can escape. No community you form, no group you build, dodges the fallout when someone decides to set fire to shared infrastructure.
We are moving into the dark forest era of the information economy. As models improve, inference costs drop, and capacity increases, the primary organism creating content online will be the bot.
Instead of building communities of people, build collections based on rules of engagement. Participants - be it bots or humans - must follow proscribed rules of conflict and debate.
That way it doesn’t matter if you are talking to a machine or a person. All that matters is that the rules were followed.
You’ll probably get one of three outcomes: regulatory capture by monopolies, self dealing by bureaucrats to enrich themselves or gain power, or regulatory capture by self absorbed ideologues who halt all progress or force it down some ideologically approved path.
In none of those scenarios is anything aligned with the best interest of the people.
Haha yes. I interacted with a bank one. It was like press 5 for mortgages but with a text to speech front end.
At the end of the day the LLM can be tricked into doing anything.
Remember that the polygraph still exists, now we will be dealing with a massive portion of the decision makers will treat as artificial inteligence not in the technical sense we use, but as real inteligence, maybe even super-inteligence.
THEREFORE A COMPUTER MUST NEVER MAKE A MANAGEMENT DECISION
—IBM internal training, 1979
It took me a while to realise that the premise is saying the same thing as the reason why we have so many "Computer says no" experiences today.
The conclusion only follows if you want someone to be accountable.
If you want to avoid being accountable, computers should make all management decisions. This has nothing to do with AI other than it provides another mechanism to do that.
People saying "I'd love to help you but the computer won't let me do that" has been happening for years now.
Websites develop abusive patterns because A/B testing lets a process decide based on the goal you want, It doesn't measure the repercussions so you have made no decision to allow them.
Management read it as
A COMPUTER CAN NEVER BE HELD ACCOUNTABLE
THEREFORE THERE CAN BE NO LIABILITY IF COMPUTERS MAKE ALL MANAGEMENT DECISIONS
The pattern goes something like this:
- this development is bad
- companies will be unrestrained in their use of this development
- there will be no rules so they can do whatever they want
- we are all fucked as a result
But then...propose that we make some laws to put rules around this stuff, also known as regulations and everybody goes "whoa hold up hold up hold up...I dunno about that part."
Dear friends - America has always been this way. Study your 19th and 20th century history. Companies will exploit the shit out of us unless we put some rules in place to prevent it. Yes, that might mean making less money in the short term as regulations cause friction. But in the long term it means we can have a better and actually livable society.
(For what it's worth I'm an American and not an uppity European or Australian taking potshots from across the pond; no offense to Euros or Aussies intended, love you guys)
Payment processing, is better than it was in 2000, but still not good.
Micropayments: this is obnoxiously expensive to do.
Discovery, and discoverability: again here we have better but not good solutions (and many of the ones that were once good are enshitified).
Pricing: this is a problem everywhere, and frankly we need the law to change in a way that is pro consumer. Publishing prices, disclosure of fees, in both services and for payment processing (that 3 percent back from visa looks a lot less attractive when it's part of a 5 percent mark up).
Customer service: well there are already companies promoting models where they cut you off and send you into a black hole (google is a prime example). Good customer service will become a differentiator, and maybe a "paid for" service as well.
Lots of blaming LLMs but I think the root cause lies elsewhere, I’m not even sure whether dismissing it as “capitalism” or “profit motives” would do it justice, because in general it feels more like the world that we live in lacks humanity.
Even in a capitalist world, a company could take a stance and decide not to purposefully screw people over, but in the world that we live in instead they look for ways to better screw over people and extract more money from them. It doesn’t matter whether your customer support is handled by someone from India, a crappy telephone tree or some voice model, when the incentive is the same - to do the bare minimum for customer “support” (in practice, just getting you to fuck off). Same for handling insurance claims and “dynamic pricing” of things - it doesn’t matter whether it’s some proprietary algorithm or just an LLM making crap up when the goal is to screw you over.
Blaming “AI” for all of this would be barking up the wrong tree (without that tech they’d just find other ways), though one can definitely acknowledge that this technology provides another convenient scapegoat, same as how you can lay employees off and just say cause it’s because of AI when in actuality it’s just greed and wanting to make your books look better.
It’s already bad. I’m not looking forward to the future. These systems are terrible. It’s a future without people that they want for some reason. I’d rather deal with people incompetent, tired, annoyed people than an LLM.
Most of these annoyances are also things that existed before AI, and will continue to exist after, because consumerist capitalism. The good little obedient consumers get abused because they don't stand up for themselves. Customer service is an enfuriating maze? Yeah, because you voted with your dollars (and political indifference) to allow companies to make customer service (the thing you pay for) worse. We bring these problems on ourselves. It's pointless to complain if you aren't willing to do anything to change it. (And if you think you can't change it, there's other nations to look at, as well as the fact that you live in a democracy - for now - unlike the rest of the world)
Hell, we already have companies whose sole purpose is to manage your subscriptions for you because you're too lazy to do it yourself. You could look at this and say, man, the world is terrible! Or you could look at this and say, man, how great is my life that I can not only subscribe to a lot of things without going bankrupt, but I have extra cash left over to pay a company to manage my subscriptions?
Don't let the hedonic treadmill and complacency trick you into A) accepting a worse life, or B) convincing yourself your life is bad when it's actually better than most people's.
For most simple mainstream questions I just ask ai instead of googling shitty results.
Most of the time ai is good enough and often better than the status ante.
People do not care if it is a stupid token prediction machine as long as the job gets done.
For the "bureaucracy has royally fucked up and doesn't want to fix it", if it is something that can be fixed with money and isn't time sensitive (e.g. you need a refund rather than get the airline to actually provide you the ticket you already paid for and want to fly this weekend): In countries that have effective small claims courts, these can be a surprisingly convenient (less hassle than the "talk to the bot" wall of the company!) to resolve this kind of issue.
I hope that these resolution methods become more common - I think the tools to fight enshittification often already exist, we just don't use them enough. A welcome side effect would, of course, be that this would impose a real cost on the enshittifiers, creating an incentive to provide proper support.
LLM when it came out, was perfect as an interface between a system and a normal human.
So many people call customer support for issues they could in theory fix themselves. If that LLM system can understand me well enough, its an okay interface.
In worst case you have to escalate anyway. My mum actually told me that she talked to some AI.
And yes normal systems are also not correct often enough. With AI/LLM software will get cheaper which should incresase quality overall.
I dont think ai/llm in this case will change anything.
Relevant change will happen due to the fact that humans can be replaced by AI/LLMs. It was not even imaginable a few years back how a good ai system would even look like. Translaters lost their jobs, basic arists lost their jobs. Small contracts for basic things are gone. The restaurant poster no one cares? AI. The website translation for some small business? no one cares.
I’ve had the same thoughts, but if you look deeper, it all circles back to what we already had: (open, transparent) public institutions, society, and government by the people. The foundation wasn't the problem; the environment was.
Along the way, social media noise, engagement-optimisation and Kardashian-style "entertainment news" infecting real news made an attention economy where, no matter how scandalous you are, attention can be minted into dollars. That is what polluted our infosphere and lead to the lack of trust.
Now, nobody trusts these previously mentioned public entities any more - sometimes due to state-actor or ad-tech disinformation, and sometimes for good reason like when the poisoned public allowed these 80s-style telemarketer-style political weirdos and their cronies to take over public administration.
Hate to break it to you but it's always been this way, and it was easier in the past when information was so much more expensive to distribute.
My agent will be in touch with yours, I guess.
> Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task
> https://arxiv.org/abs/2506.08872
> Cognitive activity scaled down in relation to external tool use. …
> Self-reported ownership of essays was the lowest in the LLM group and the highest in the Brain-only group. LLM users also struggled to accurately quote their own work. While LLMs offer immediate convenience, our findings highlight potential cognitive costs. Over four months, LLM users consistently underperformed at neural, linguistic, and behavioral levels.
I also enjoy the series. But sometimes my friends send me things and I'm like, "not gonna read all of that."
Just because you're friends don't want to invest the same amount of time that you want to invest in your own personal enrichment doesn't mean they're getting stupid.
Not meant to be snarky. It's been two decades now since my first wide-eyed entry into the workforce, moving for new opportunities, meeting new people. it's been great. There's a lot of smart people out there. I also realize that many people I seen as smart had more access to more content then i did. i still appreciated their sharing , it was enlightening to me. But after 20 years, I think back and it's literally quoting things from smart youtube videos. and regurgitating the latest thought leaders.
We all do this, but like you, what's meaningful to me is the chewing, the dissection and synthesis. coming together to share different perspectives and so on. i've had those friends too! it's just not 1:1
I read one of his last week? and didn't like it that much. I read this one despite it because its quite high on hn for whatever reason.
I don't think everything is lies and i don't like how he thinks a LLM is just some bullshit machine.
Its also waaaay to early to even understand were this is going. We as humans have never had that much compute and used it this particular way. It could literlay be the road to a utopia or dystopia. But its very crazy to experience it.
His article series feels so negative and dismissive, that i'm not taking anything from it.
There is so much more research, money and compute behind this AI topic right now, every week or two weeks something relevant better/new comes out of this. From 2d, 3d models, new LLM versions, smaller LLms, faster inferencing (Nvidias Nemotron), we don't know how this will continue.
And the weird thing is that he clearly knows plenty about LLMs but it feels so negative dismissive, hard to put a finger to it.
We intuitively know that an employee will be punished and may get fired if we trick them. Many of us won't try to trick human employees as a result, because we would feel bad if they had bad consequences as a result of our trickery
There is likely no such hesitation around tricking LLMs. I know I personally wouldn't feel bad about it at all. Mostly because any computerized customer service process is annoying so anything I can do to limit my time dealing with it is a win in my books
I haven't a had a good experience since AT&T bought my previous ISP and forced me to switch to a different subsidiary.
It is an incredibly vexing situation to see whatever you're an AI hater or enthusiast.
I, for one, welcome our trillion parameters multiple layers overlords.
Rather than dismissive, I see it as effort intensive. The conclusions can be negative, but they've spawned so much discussion which i think is great.
The average person on the street is familiar with consumer-facing AI but doesn't think it's really alive/magic/the solution to everything. Our supposed best-and-brightest are the ones flogging the horse.
The future of everything is lies... sure, well so was the past, what's your point?
When we say "snake oil salesmen" we literally had people selling these mineral oils, fraudulently, for a century. People yearn for a time when there wasn't an antagonistic relationship between buyer and seller, but it has never existed. There is only one way have that relationship, and it's by having a personal -- repeatable -- relationship with your seller, and that's usually expensive, even if that expense is not monetary. It just the game theory of repeated games.
The easiest way to establish this behavior is to be a regular at a bar while being a good patron that creates a positive experience for everyone else. There are multiple places I frequent where I have no doubt in my mind that if there was an emergency, they might ask me to help out in some way because I'm trustworthy. And they could hand me an envelope of money, knowing I'd not steal it, because it's not worth it for me to take a bit of money and ruin my welcome at this place.
That's not going to happen with online sellers, and it's not going to happen with most corporations. The promise was always "save money through economies of scale, such that you still win even though there is no relationship." For the most part, that's proven true. But as everything has been outsourced to China, and anyone can make a basic version of everything, we're running up against the limits of those benefits. And I truly believe that brands are going to start mattering a lot more going forward for marketplaces.
Shop at Costco because they care about their customers. In other areas where companies don't, you're going to have to do the legwork to find out who is going to treat you right. That sucks, and it'll probably be more expensive, but that's the antagonistic relationship that buyers and sellers have naturally. Regulation helps, yes, but it's not some cure all. The expectation that you can just go with the person with the lowest price and get quality services is not a thing that's every existed.
The real issue is new amendments are needed. But that’s hard. You need 75% of states ratifying. And that seems impossible today on any topic.
A computer can never be held accountable. Therefore, since all management decisions must have accountability, a computer must never make them.
Incentives make the world go round, so even if people recognize the issue, they would rather it become someone else’s problem, than willingly harm their own future.
This does not matter without antitrust, which is why customer service became bad in the first place. 30 years ago, the low quality of customer service we complain about now simply didn't exist, at any size or professional level of business, and never had.
If a company back then had the customer service of the average company now, or even the average government agency now, people would have suspected that it was a covert front for criminals or spies.
If a company doesn't have to compete, it can cut everything until it only has the ghost of a product and a billing department. You don't boycott monopolies, monopolies boycott you. If three companies put you on a list to not have internet, phone service, a bank account or a credit card, etc., you just can't have them. You've become a European human rights judge.
It would be great if people chose not to do business with the former, but many simply do not care. They may think only other people get screwed. They may not take the time to think about it, especially if the company spends a ton of money obfuscating their misbehavior. Quite a few actively defend the right of companies to screw them.
Technology multiplies that like a lever. We weren't prepared for capitalism before LLMs and we're massively under-prepared now.
Not a fictitious example.
Also, if the effective tools do exist, count on American companies using the American bribery-based political system to change the laws to dull those tools or to eliminate them.
Again, if you live in some consumer-friendly country good for you, I’m just saying how it plays out in this one society. I’ll stipulate that it’s all our fault blah blah blah.
Every time I call the Costco pharmacy, I just hit 0 immediately because: Phone. Trees. Suck. They have always sucked, it's just an awful, grindingly slow way to accomplish ANYTHING, and it's so, so much easier to, when I need help, get a person on the line who can figure out what's gone wrong and sort it.
The only people benefiting from cutting that down are the scum class (combo of shareholders and executives) and who's shocked, really. Everything is being ruined nearly at all times to benefit the scum class.
(FYI, I didn't downvote your comment)
Maybe it's something like that AI allows them to indulge in their shallowness/laziness by giving them the impression that they're not doing that.
> People are very excited about “agentic commerce”. Agentic commerce means handing your credit card to a Large Language Model, giving it access to the Internet, telling it to buy something, and calling it in a loop until something exciting happens.
I think you're confusing this for the other side of things. The article talks about how some people already use OpenClaw and the variations, give them access to bunch of stuff including cards to purchase things (sometimes virtual and limited cards), I think that's what the article talks about when they say "agentic commerce".
Obviously a intentional simplification in the language the author uses, but I think it gets the point across at least.
Of course, what those who call for regulation of AI or other nascent technologies are really saying is, "Unqualified and/or biased and/or corrupt and/or dull-witted people should make decisions that affect us all, based on incomplete, misleading or rapidly-evolving information, with the power to enforce them at gunpoint."
And then they wonder why other countries beat their own.
The company that deployed the LLM is lying to you. The people who made that decision are the ones who are culpable.
We both agree that it’s terrible.
I think it’s important to have an enforcement mechanism to force companies to do what they are responsible for doing. An Anti-Kafka Law, so to speak.
That is not known. Things could easily just get worse, and IMO that is far more likely. Every civilisation has collapsed, ours is clearly in decline, and AI could likely accelerate that decline.
I'm not exactly seeing the progress here. AI helps us write some software a bit faster? Doesn't seem revolutionary to me. Is it having any significant impact on peoples lives other than the various economic forces? I'm not seeing it.
> Yeah, because you voted with your dollars
In our system? No. In crony capitalism the companies who win do it through manipulating the political system. And when the government inflates the currency and destroys people buying power they simply cannot afford to "vote with their dollars". This is hilariously naive.
As the author said:
>I suspect that like the job market, everyone will wind up paying massive “AI” companies to manage the drudgery they created.
Statements like this make me feel like I live in a different universe with a different implementation of LLMs than other internet commenters.
You have no argument here. Make an argument then we can talk. Right now it’s going in circles.
AI isn't like that. One problem is that it's rather generally misunderstood at this point. "AI" is not "intelligence". It's intelligence-adjacent, and something like LLMs is part of our psyche...the subconscious facility that allows us to form sentences without really thinking about it.
At any rate, I have to agree with most of the points the blog author brings up.
It’s not LLMs replacing workers, it’s people. People who have a lot of money and don’t sell their labour for a paycheque. And the systems that compel them to such actions.
This is the definition of doomerism. Of all the ways each society has adapted to industrializaion, of all the technologies we've adapted to, and our continued use of all of them despite any negative impacts, you think an LLM is going to end things? An overcomplicated autocomplete?
There is no evidence whatsoever that suggests our society would collapse under AI. An imagination is not evidence. I mean, it's just a tool! Like the car, the telephone, the genetically engineered seed. Humans always freak out when they see something they don't understand. But it turns out there's no tool we have created that we can't control. We're stupid, but we're not idiots.
Also, I’m reading this comment thread instead of TFA because I didn’t find the previous part I read that great. And I’m not an AI proponent, more of an AI skeptic.
It could lead to significant progress in every single research area.
I'm at least very impressed about the amount of open models and that it doesn't hold up that the gap between public and private diverges massivly. Public is probably one year behind.
My point with my mum should imply that it was successful but for sure at least you are pointing something out and now we can talk about it: My mum talked to an AI and it helped her.
And it's a one day old account.
I mean, did you not write with ChatGPT and were surprised how well it response?
I'm schocked how well i can talk to an AI through some app like Gemini or ChatGTP. A few years ago i couldn't imagine building such a generic system which such high quality of understanding.
I was playing around with dragon naturally speaking and similiar dictation tools 10 years ago and it was horrible. And that software is expensive.
If you look how normal people use a computer, they are slow just because they don't understand basic drag and drop. Or they are unable to just create some java or php script to convert some data or clean up some data. I would just write a php script reading some csv file and converting stuff around and was faster than everyone around me.
Tool calling is bonkers.
And i tried to break GPT-3, i can literaly write an english sentence and just dropin german words, it was already that good.
Its often enough shitty in doing exactly what i want, but the quality is massive to everything we had before. Massive.
So my main concern here is that my experience may be a microcosm of the shallowing of discussions correlated with some people's increased use of AI. That worries me.
It's more of a meta point to me. I get that this series isn't landing for some people, yourself included, but the meta-observation is that given something of roughly equal substantiveness as before, these friends' motivations for long form content and discussion seem to have atrophied, perhaps largely due to the addition of the AI summary reality cipher to their lives.
Of course, correlation isn't causation. Maybe they both just got older and more lazy, but given their reliance on AI summaries in other debates happening recently, I'm worried.
An LLM can enhance the interface of a system and can be really useful in that despite its imperfections. But that’s a very different claim.
How many people can change the time on their microwave?
How many people can ask an LLM through voice or text to change the time of the microwave?
A LLM is an interface to a service if you add a MCP Server. Now i can ask Jira things like "hey whats my current task? And what do i need to do?"
Its also an interface to documentation. I asked it to help me build up a hugo templating based website because just reading the hugo docs did not help me as much as the LLM did (and that was 2 years ago).
In best case, as long as an LLM is not AGI or ASI, we have good tools with validation behind the LLMs before the LLM becomes the system itself.
What about configuring your Jira views, and then bookmark the resulting URL with a nice name like "Jira: Tasks in Progress" or "Jira: Important Tickets". That would be way faster than any LLM prompting.
> Its also an interface to documentation. I asked it to help me build up a hugo templating based website because just reading the hugo docs did not help me as much as the LLM did (and that was 2 years ago).
Those kind of claims would be better if the person has written down the goals before the activity and then score the end result according to those goals. A lot of time, there's a lot of post-rationalization (like "I spent time on it so the result must be good"), especially from non-expert.
Only if you care about doing things fast.