Please stop the AI confidence theater
elenaverna.com
elenaverna.com
You have, in fact, just engaged in it in your comment by offering an opinion in hopes that others will read and adopt that opinion. In fact, you posted your comment on the marketing website of a well known private equity firm.
Modern marketing frequently "rides the line" of what's legal and has little to no concern for ethics. Teach kids how to annoy their parents so they buy your toys? Sure![1] Prey on teenage girls' insecurities to sell them cosmetics? Of course![2] Lie to folks that they're going to "win big" with your gambling app? Why the hell not!
[1] https://en.wikipedia.org/wiki/Pester_power
[2] https://en.wikipedia.org/wiki/Effects_of_advertising_on_teen...
But that's not really what the problem is, since marketing would still be annoying even with that solved. The problem is much more akin to a tragedy of the commons type of thing, where, when everyone is shouting, you can't hear what you want to hear. Usually, that's the thing that bothers people. Billboards, unsolicted emails, etc.
I think for people who really get annoyed at marketing it helps to take a couple of steps back and look at how we fit into the natural world, everything that frustrates you about human society is something you'll find an analog for.
Things like brightly colored frogs that are poisonous, fur spots that look like eyeballs looking backwards, peacocks, bugs that looks like sticks, etc.
All of these are what's called emergent behavior, and they're intrinsic to all complex adaptive systems, like natural ecosystems, and human society itself.
To address your point - the flowery and innocent core promise of the field may have been true when it was first invented thousands of years ago. Exposing people who actually want to learn about goods and services to your offering in an attempt to establish a mutually productive relationship (they get something they want, you get money) is fine. But that stops short at about the 0.01% mark on the way to the beast that marketing became today. Modern marketing is about pure value extraction at any cost. Modern research and tech has enabled them to find loopholes in every relevant regulation, flood every empty crack of the internet with garbage if it means they get an extra cent out of it, study the flaws of the human mind to discover the best ways to abuse it into buying their thing, plaster every object in existence with screaming ads and audiovisual trash to force people to internalize their message. There is no natural cap on marketing, so we've long since moved past unobtrusive, good-natured promotion into full-blown insanity.
I think billboards should be completely banned, or at most relegated to designated districts of dense urban areas. I'm in favor of banning products that are addictive or harmful to people, like gambling, tobacco, and the like. I think that conflicts of interest should be disclosed, and false claims should have consequences.
But I think it's helpful to start the discussion by recognizing that the state that we're in now is the natural state of affairs. It's not evil, and we all engage in elements of it most of our waking lives. But unchecked marketing and persuasion will flood every crack of everything if you let it. Which is why you have to decide what you are willing to permit as a society.
It has to be done coercively, though. The incentives are so unbelievably strong to cheat that shame or persuasion isn't a useful counterweight.
A lot of outrage in marketing comes when it stands in the way of your product experience. Think TV ads that interrupt. Or fake ratings that abuse a platform.
i hate marketing. but most everyday, 9-5 in a cubicle marketers are just trying to feed their families and keep a roof over their head. its just a job, it doesnt need to be an identity.
or, you know, people can encourage all marketers to commit suicide a la bill hicks. its fun to be edgy sometimes.
(edit: i find it so strange that suggesting a tiny bit of empathy for someone grinding away in a cubicle is so objectionable to the people on the venture capital marketing forum, but alas, i have much to learn about the quirks of HN)
Unless you are severed, of course.
Of course not, because an identity is just an idea, a word. You lose sleep, or don't, over what you actually do. And you cannot do a job without doing it, by definition. Some jobs are kinda scummy, some scummy people enjoy them
I remember over 20 years ago when at a party this other kid suddenly started talking about his job in a call center, selling lottery tickets to grannies basically, and he grinned and told me he could imagine I could be good at that, that the money is good if you hit the quota and whatnot. I thought, what do I care, I'm also good at running real fast, so does that mean I snatch purses now? No, because I do not enjoy that, and would lose sleep over it.
> or, you know, people can encourage all marketers to commit suicide
First off, there is a lot of grey area between "they're all poor victims" and "they're all demons". Secondly, why "all"? You're basically using the good people and poor saps to cover for the bad guys by waving your hands around a little.
Saying that it’s “evil” and wondering “how they sleep at night” is unfair to your own intelligence and its ability assess the value of this function.
I don’t like advertising, but I think calling the act of or working in marketing evil is really reductive vilification. Try to take their point of view, and realize almost nobody is trying to make the world a worse place.
Yeah, this is just absurd. From the other side, probably marketers wonder how anyone implements these incredibly user hostile interfaces to make it hard for people to cancel, and how anyone could work on products such as Google and Facebook (even as they use them for work).
It's easy to scapegoat a group you're not part of, but it's generally not correct. Things are more complicated than you imagine.
I guess if "Being good enough that people recommend it to others" is marketing, then yeah I suppose I'd agree. But when people talk "marketing", they generally refer to outbound one, and we could easily live without that and still find products, just like we did before "marketing" was the pest it's identified as today.
With that said, everyone believes they're a good person, every single individual on the planet probably, within their reality and their perspective. Including marketers, unbelievable as it sounds.
They sleep just fine. Probably because they believe they do just a small part that's not harmful by itself and if they were not doing it it would be someone else.
Why single out sex workers here? I understand the other examples, but unless you're talking about sex workers tricking people/robbing them or something else, it seems like a service-profession like any other, like a cleaner or what not, but I'm guessing you wouldn't actually put 'Actual mercenaries and cleaners kinda fall under the “all is fair in love and war” frame'.
I think a mix match of perverse incentives, rent-seeking, cobra effect, regulatory capture and institutional corruption all together lead to the inevitable situation of institutions or professions whose incentives depend on the persistence (or even worsening of) the problem it claims to solve.
Of course I didn’t “single out” anyone, but you got that.
Sure there will always be subjective differences , but a lot of negative impacts on this planet can be attributed to the distortion of actual merit and the distorted proportion of what should be correctly attributed to one individual .
outreach, discount sales, testimonials, business pages are all marketing. "purest form of evil" unless you're doing it i guess.
(or, perhaps, there's some nuance to be had instead of proclaiming marketing in general to be one of the "purest forms of evil in existence"...? )
The idea that you can’t even in passing mention something you spend all day every day working on, because then it’s “marketing” is exactly the sort of consequence which leads to so many of us loathing marketing as an industry.
There’s plenty of marketing out there that just tries to make information about a product and service available without focusing on driving home higher revenue at any cost. That’s usually advertising, not marketing though, but it does exist.
https://genius.com/Bill-hicks-on-advertisers-and-marketing-a...
It's awkward because you can feel how much he means every word. Of course it's part of the act and you're supposed to find it funny, but at the same time, he very much means what he's saying.
Sadly, he was never able to rid the world of all these fevered egos tainting our collective unconscious and making us pay a higher psychic price than we imagine.
Ironically Bill smoked like a chimney and for that we can also largely thank marketers.
But since the last year or so, I can't watch him anymore. He sold his channel (and his brand, literally himself) to some kind of YouTube content company and the videos he puts out now are just not watchable. From what I can tell, he mainly does only the presentation now, with only a minor amount of editorializing. Other people seem to do everything else. The visualizations are impressive but the video title/thumbnails are pure click-bait (to the point of being factually false), the videos are WAY longer than they need to be, and he'll repeat the SAME information multiple times just to stretch the time out to 45 minutes to an hour.
I like a good story, but it's really hard to pad out most disaster videos into that amount of time unless you have something more to offer than say, "well, the official crash investigation said this and that." His videos now feel a lot more like those old Discovery Channel documentaries that were basically surface-level filler content in between the ads.
how many times have we all seen marketing departments or sales departments in our companies entirely misrepresent the abilities or purpose of a product we built?
it’s fucking unreal how many times i’ve seen on here where the engineers of a product were like “don’t blame us, our team was screaming trying to be heard that the marketing/sales departments are outright lying about the capabilities.”
at which point they’ve often twisted and bastardized the product opposite of the reasons we built it to begin with.
Trigger warning. Could not finish that sentence.
1) cool people do cool things, start a scene
2) chill people who enjoy watching cool people do cool things spread the word
3) posers get wind and show up, the scene loses its vibe but reaches critical mass
4) advertisers show up looking to monetize the scene, driving out the cool and chill people who are allergic to advertisers.
5) the scene is now dead, filled with posers and ad execs
It’s the culture of the org and that decides how the ammo is used. What you want is good devs AND good marketers building and promoting great product.
Anyone in any role who works at facebook holds culpability. Marketers aren’t building all those user hostile features.
With printed books for web development you want a recent book. I am sure there is much to be learned from the 2002 book, but I want 2024+.
The list of titles starts out strong, with titles such as 'Web development with XYZ' but the 2024+ titles are 'web development with AI and XYZ'. Which is probably jolly interesting, but I want the fundamentals of XYZ, not AI + XYZ.
Dunning Kruger springs to mind.
Some of the 'egging of the pudding' is most interesting, I have a friend in scientific publishing, and, with the yearly performance review and 'strategy meetings', the friend, who manages a vast department, said to the boss how the plan was to go all-in on AI. This was music to his ears! The performance review went extremely well, the right things were said and yet nothing specific was committed to, just this smearing of AI everywhere.
Does this friend or the boss, or the team, have a clue what they are going to use this magic AI for, or what the results will be? Who cares, bossman can now present 'his' AI strategy to the board, with the press release going out and the share price going up.
This was almost a year ago and I daren't ask about how the AI thing is going for them. I suspect the ship is still sinking (open publishing is eating the industry) and that the magic band aid that is AI might not be working for them.
After coming away empty handed from my book search on O'Reilly, neither wanting an out-of-date book or an AI-centric recent title, I am wondering where this is going. Presumably at O'Reilly they also decided to go all in. Maybe there was a manager like my friend, telling their boss over-confidently that this was to be the strategy.
In tech we are always dealing with unknown unknowns. AI just makes it easy to gloss over this, meaning that we have a lot of Dunning Kruger going on. The further up the management chain one goes, the more Dunning Kruger there is.
It’s better starting from an old books and retrace the updates from that. And with the benefits of hindsight, you can get truly good ones cheaply.
But now you would have 'how to do grid layout with Claude' or something equally daft. And you can bet your bottom dollar that the book would itself be written with AI.
IE6 hacks are skills. The real understanding is how CSS helps with layout. Grid may not be in an old book, but the old will give you enough foundation to easily grasp it. So something like “CSS in Depth”[0] (which talks about grid BTW) is a good candidate to buy.
I think you can get good materials if you focus on getting books that talks about a system (design, theory, implementation), not techniques(skills, how to). Systems evolve and are build on top of each other, or influence each other. Skills come and go.
It may make pity scams less effective though so I'd expect the implementation to change to focus more on creating urgency and less on altruism.
The prospect of not being able to pay for food or rent carries a desperation that isn't there when just wanting to buy a video game or hire a Hooker :)
Maybe with hard drug addiction, but that's a different societal problem to tackle.
> it just re-baselines
If fewer people are chronically -desperate- for money, it could leave a smaller workforce pool available to professional scammers for "hiring ghost writers to pump out low quality ebooks" or "paying people to leave 5 star reviews" etc
A lot of people I know are forced to use AI at work. They universally tell me that their coworkers generate awful PRs with bugs, nonsensical code and fake unit tests. But they also universally tell me that they are different, have special workflows and prompts that create good code.
The psychology behind this is obvious. Hype and the literal threat of being fired forces everyone to develop coping mechanisms. Bragging about your own adaptability is one of those mechanisms. Unfortunately, the scale and intensity of this and the fact that the mechanism is clearly weaponized for marketing means we're living in an increasingly deranged society.
Also at work: the Dunning-Kruger hypothesis, Gell-Mann Amnesia and motivated reasoning.
This is social collapse.
The truth is there are prolific developers like Antirez who have built quality new projects at an incredible pace (Dwarfstar 4, Redis features).
But as unpopular as it is to say it, in the working world ~80% of developers pre-AI mostly just attended meetings, did a little busywork and committed small patches here and there. Probably around 20% really moved the needle and contributed the bulk of net new code.
Those 80% were constrained in the volume they could output pre-AI, but now they are unleashed to do a large amount of net new work but many without the skills to structure it well+maintainably.
It doesn't help that most management has been pushing on LoC over quality the past year.
I truly believe most companies as they exist today are not structured for AI. The amount of technical debt that will be created at a rapid pace is basically time delayed self destruction for most codebases if you let people run amok with low contribution standards and rubber stamped approvals.
If you treat each AI output as a small well-scoped, well-tested module, which interoperate with each other through well designed APIs, you can have high confidence in quality. But majority of people are pseudo-vibe coding and creating spaghetti monster codebases, and there's really no way to stop it without strong and tight technical oversight.
I like to imagine a similar dynamic happening when previous transformative technologies were invented: when the power tools like the chainsaw were invented, I wonder if there was a cohort of carpenters that dismissed them because of how much damage they could do in unskilled hands.
You could say that AI turned them from stupid and lazy to the famously dangerous combination of stupid and hardworking.
There is a lot of dumb money chasing the AI dream. That money is not asking hard questions about return on investment yet. But, the narrative is starting to shift as evidenced by this article. Even Meta is questioning the value of AI Agents.
In response, we see Anthropic, OpenAI, and Microsoft creating giant teams of FDEs (forward deployed engineers). The idea is to keep the AI dream alive one way or another because once that dies, the whole thing comes crashing down. So, dumb money is going to put up a strong fight in the marketing-sphere, and articles like these are needed to counter that.
If someone told you this (their coworkers generate bad PRs, but they generate good PRs) in the age before LLMs would you have also declared it psychosis?
Having to deal with lazy coworkers who submit bad PRs has been a feature of workplaces since the dawn of programming. Programming languages are a tool and they can be used or misused by the operator.
The difference now is that a lazy developer with an LLM can become prolific with their sloppy output and it’s harder for a lazy manager to notice.
If you’re trying to imply that nobody can use LLMs to good effect then that’s just denial at this point. The way good developers use AI isn’t to prompt and then submit PRs. It’s used as a pair programming partner. The developer still writes, reads, edits, and is responsible for the code.
You think this a clever gotcha but it's not because no one was generating PRs before LLMs.
My colleagues are absolute work horses. They work for 8-10 hours minimum. Always responding to emails, putting out fires. Always typing furiously, producing lots of code. You are waiting for it, I know it and yes, it's all shit and all those fires are of their own making.
I picked up stuff in a month or two that took my colleagues years and I'm not a genius, as you can undoubtedly tell. I am just exceedingly lazy and I honestly think it's pathological. It got me in a lot of trouble in the past, but now, with the way AI is going I'm having a blast.
I have some proclivity for architecture and directing large complex projects. I hate typing code, but I love pointing out how it should be done. This shit is a life saver but for the love of God if you are not supremely lazy and averse to code, don't use it. My productive colleagues + AI = goddamn disaster.
i dont think there is any software on the planet that i would consider "truly life changing", so i find it a bit weird to hold ai up to that standard.
as a note, i found this particularly funny:
"It’s doing more harm than good." followed immediately by "This post is sponsored by Firecrawl. Firecrawl is the web data API to search, scrape, and interact with the web at scale."
I think the author made this point because earlier they mention how people tell them AI changed their life.
> So I started asking a simple question whenever someone told me AI had changed their life: ‘Cool. Show me.’
yeah but people say that in casual conversation all the time. my wife, not long ago, said her new facewash changed her life. its a figure of speech.
Every week there is some new method to make LLMs produce code, ranging from Langchain, MCP, Agents, Ralph loops, gas town, just loops.
You are stating its hyperbole, when people are walking around without closing their laptops, or are awake all night running agents.
Every software purchase process I have been involved in I have asked the vendor to demonstrate their claims against our requirements.
With AI, I am expected to defend myself from claims that my requirements are wrong.
For context
Can we PUHLEASE stop this AI Confidence Theater, people? It’s doing more harm than good.
Web browsers truly changed the lives of giant majority of people on this planet.
I don't think there is any software on the planet that has accumulated 1.5 trillion dollars of otherwise-useful money!
I think the combination of web server and web browser comes close.
Surely you're kidding right?
You can have a medical emergency while sitting on the can in a bathroom, and then call up your doctor who can magically see exactly whats wrong as if they are looking through a literal crystal ball portal, and then they can immediately search through a corpus of billions of medical papers to find the best solution, and relay that back to you. Then when the call ends, you can summon a car to come drive you home like a magic carpet, and the dude driving it gets paid automatically with coins that you weren't even carrying at the time.
Magic is real. We are f***ing wizards.
You seem a little full of yourself. (That’s putting it charitably. You seem full of shit.)
I consider LLMs life changing in the sense that the Internet was life changing: it makes information much more accessible. In the olden days, learning just about anything outside of your immediate circle (family, friends, teachers) meant a trip to the library or bookstore. If your local library or bookstore didn't have it, you were SOL.
The 21st century problem is different: too much information, while too much of the accessible information is repetitive and of dubious quality. LLMs are fairly good at summarizing human knowledge. If your research is important you can ask the LLM for targeted sources to: vet the LLM's summary, vet the source of the summary, or get further information.
I think hyperbole is problem with the "life changing" crowd. Too many people expect the LLM to do the work for them. Even something like extracting information from a document is your work, not the LLMs work. Writing a piece of software is your work, not the LLMs work. Anything where your responsible for the outcome and where assessing the outcome would involve reproducing the work of the LLM not going to be life changing because it means you still have to do your job.
Leave computers to do what they are good at: massive amounts of calculating and collating. Doing jobs that are beyond human reach because we are not particularly fast nor tireless. Doing jobs where it is more efficient to throw a machine at the task than it is to organize armies of people to do the same. In that respect, LLMs are just tools. As tools, LLMs aren't terribly different from the original computers.
FreeCAD genuinely changed my life: it made it possible for me to do things I was sure I would not be able to do, think in ways I never though I'd grasp, visualise and design physical things I thought required expertise I couldn't ever develop, and develop that knowledge, and it does so in a way nobody can take away from me.
LLMs, at best, feel like a bespoke, error-prone reference library combined with an addictive drug that I could get some value from, but risks stealing the joy from everything I do.
Maybe a bait or not, but quite important for planes, for example. Or just cardiac pacemaker...
- My life was different before Skype allowed my business to go fully remote in 2012
- Stuck in Argentina during one of their troubled periods, AirBnB was the only way we could pay for accomodations because you couldn't withdraw USD or EUR, nobody would take a credit card, and it was a mess all around
- WhatsApp (which I don't use anymore, now replaced with Signal) changed the life of my family and how often we communicate with each other (and meet in person)
- My health and fitness is unrecognizable since I started using a couple of FOSS apps for calorie tracking and workout progressions (Waistline, GymRoutines, Podometre among others)
- Of course my life is different due to my own software because I absolutely hated editing videos myself, and OneTake has removed one of my top weekly pains :)
That's just off the top of my head.
I think the whole point of creating software is to aim for it to be life-changing for someone. It doesn't have to be world-changing. But if it's life-changing for no one, why bother?
Off the to of my head web browsers, the algorithms that made WiFi / 5G / Bluetooth possible, public key encryption, GPS, SQL, TCP/IP, Reed Solomon codes spring to mind. Without them like today would be very different. Sounds life changing to me.
The current valuations of AI companies suggest AI will have a similar effect. I think it will, in the long term. People tend to overestimate/oversell the short term effects of technology, and underestimate the long term effects.
Now, four years later: I've finally been able to purchase a 5070Ti (from a ~1080Ti~ – so, an upgrade). The past few days have been overwhelming as I explore the rights/wrongs of bigger offline models (more parameters, even if They're [only] Made Out of Weights).
----
My twin, an actual brilliant engineer (me: amateurhour), after playing with this new "toy" [discussing chessboard layout, electrical engineering concepts] deduced (accurately IMHO): "it's able to be wrong FASTER – don't let this be discouraging it's an incredible piece of hardware but I still cannot trust it for citing reference material." My general Qwen3:14b model was unable to troubleshoot a bespoke coding issue (as expected, it's not a code-er).
I have minimal coding experience, but he is a professional hardware engineer; I've been self-tasked to play around with "any coding model that might solve a bug I'm having with a current piece of firmware" – so that should be an interesting conceptual experience for both me and twin.
Anybody have a good coding model (it's whatever language Atmel Arduino uses, by default) that runs on a 5070Ti, best?
Atmel Arduino is mostly assembly (rarely used and assembly is generic to everything) and C. But you can’t use the usual C libraries because they are tailored to OS like Linux or Windows and the hardware those run on (i386, amd64, arm,…). That’s the primary constraint you need to implement in whatever LLM tooling you’ll be using. The majority C examples are about Linux or Windows API, not the simpler AVR standard library (the gnu one).
Other than that, on Linux and Windows, the hardware is abstracted away. For Linux, You don’t interact with the keyboard and mouse, you use libinput/x11. You don’t interact with soundcards, you use alsa/pulseaudio/pipewire,…. With AVR devices, most abstractions are only one layer deep and mostly help to avoid dealing with the boilerplate of bus protocols. You spend more time with datasheets and diagrams than with code.
Start by asking the model why "whatever language atmel arduino uses" makes no sense.
Atmel is a company that makes chips, some of them are AVR some of them are ARM. The language? well, you can get a compiler and use (mostly) whatever you want, unless you are talking about the machine code.
Arduino is the software platform + dev/prototype board manufacturer.
However, what I meant was: whatever the default programming language used when running Arduino Software Suite.
Which is "essentially C++" according to my LLM query.
You will find this most prescient: https://news.ycombinator.com/item?id=48557728
<https://www.techpowerup.com/gpu-specs/geforce-rtx-5070-ti.c4...>
https://www.techpowerup.com/gpu-specs/geforce-rtx-5070.c4218
I really wish NVIDIA would name their products better (e.g. the 5070Ti is closer to a 5080 than a 5070, in performance).
[1] Just as examples: 136 tok/s Llama3.1:8b; 65 tok/s Qwen3.5:9B; PoExile 4k@144Hz... it can do each while the majority of the CPU cores mine Monero =P
If only journalists would ask similar questions to corporations. "Ok, you made your entire engineering department agentic. Show me how exactly that contributed positively to your bottom line"
While the coding horse has been beat within an inch of its life already, I'd recommend throwing Codex on 5.5 high thinking with Computer Use + auto approve at the next thing you're about to spend 5+ minutes on to start to get a feel for how well it handles a broad range of work across literally any surface you interact with today. Use voice mode & mobile app for remote control to seriously watch the friction break down.
Is it always perfect? Maybe not - but for a dramatically increasingly slate of tasks it's becoming a no brainer to offload the busywork and raise the bar on what a single person can do.
It's natural to have hype when you see where this already and where it's going.
An excellent epithet for people who depend on AI!
I was told growing up "you won't always have a calculator in your pocket" and yet now my phone has an offline LLM on it.
But does your world stop when the phone is out of power? And does every task require a roundtrip to the internet?
Are you suggesting that losing electricity in the modern age (entirely absent AI) doesn't upend one's world?
You seem to be saying "we should avoid this thing because we'll become dependent on it," but we're highly dependent on all manners of technology for all sorts of things and would seem to be better for it.
While I think the person you are responding to has made a low quality comment, I will say that it is very, very revealing that so many AI advocates actually seem proud of their absence of basic math skills.
Most of the skepticism I encounter on this front is due to lack of proper direction, process involving planning and review before execution, and appropriate attention given to evaluation and feedback loops.
If you asked the smartest person in the world to YOLO a task with the sort of instruction the average denier uses to evaluate an LLM, you'd likely find they wouldn't get back what they were expecting either - and if you're evaluating on subpar models/tools, you shouldn't be surprised to get subpar results.
I use LLMs all the time to help me diagnose bugs and work through my designs, but again and again, I am super unimpressed by their coding abilities. I can see how in some cases with a proper harness they probably do a decent job at certain tasks, but almost everything I try to do, they flail.
If you're just saying different tools are best suited for different problems, apologies - that's my take as well.
This stuff is moving fast, and if you aren't evaluating SoTA on at least a quarterly basis, you're going to have a bad time.
The upside is that we’re competing against these folks. Let them continue to stick their heads in the sand.
You two are obviously smart people, but have you considered that maybe you are unconsciously being backed into a position that is just as absolutist as those you are arguing against? That you are just taking the opposite side in a binary argument where each side sees only one side of the coin?
I also wouldn’t characterize my position as absolutist or one side of a binary. Interesting that you’d assume that.
Or even worse: https://news.ycombinator.com/item?id=48775860
I have no idea what stuff like "is it always perfect?" means because it varies so much from person to person. Too many people have different expectations, are working on different problems, or have different standards or goals for there to be a common constructive discussion.
The only way to get a sense for these systems is to use them on things you know well, and everyone knows different things at different levels.
People also tend to underestimate how fast this is moving and base their take on dated and subpar systems for a variety of reasons, a key one being that the firehose is too big for any one person to have a proper focus on all of it.
I tried having them work on a LSP. The fact I got a one shot half working autocomplete based on my existing work was cool, but again, they flailed on incredibly simple things like file path normalization / converting from a URI and I had to rewrite a decent amount of code. I don't think I saved any time
People keep throwing this out there but I keep wondering where are the receipts? I am seeing less interesting software released, anecdotally I know, since AI has taken hold, than before.
The crucial breakdown here sounds like either lack of proper context/harness or insufficiently capable model (there's a huge gulf between GPT-5.5/Opus 4.8/Fable class models and anything not from the big three) or both.
If you aren't seeing remarkable things being done with this tech, I'd argue you aren't looking hard enough. I understand there's a lot of noise obscuring the signal, but that's always the case with a "big thing."
Also this is a remarkable (and realistic) evaluation of where these systems are for general work which speaks to both the room to grow as well as the pace: https://www.remotelabor.ai/
For some practical examples of what the leading consumer grade AI can do, Ethan Mollick consistently has great writeups with demos: https://www.oneusefulthing.org/p/what-it-feels-like-to-work-...
The second link looks lab-sponsored.
The third link is a blog post of a lot of hype and little substance. (I tried fable twice and I was unimpressed)
I'd love to see an example of something someone made that is not affiliated with these corporations.
The thing is, I imagine there ARE examples, AI isn't completely useless, but the amount of signal around this vs the hype noise is out of control. My guess is 99% of what's built with AI is useful to approximately one person. When I hear AI boosters talk it sounds like there's an entire economy of billion dollar corporations I never noticed, and then I usually found out someone got excited because they prompted Claude to make a flight simulator (where the plane flies sideways)
All my examples are private which also probably applies to many other folks here. In my personal projects and friends stealth startups that are greenfield, the boost is 5-10x for coding. For my midsized enterprise employer, we have tracking of over a hundred diverse projects and the average reduction in dev years to completion is about 50%. Not 10x, but not chump change either.
Sent from my Claude Code
A new model is released, AI fans hail it as huge shift in whatever metrics the AI vendor has gamed this time, and all criticism is shrugged off as "not up to date" and met with "try the new model!" Then, once level heads actually put the claims to the test and find it wanting, criticism is met with "you're just not using it right, you have to learn how to prompt/context/loop engineer for best results" until the next model comes out and this argument repeats.
Skeptics continue to move the goalposts on what constitutes this mattering, but the fact that frontier systems are making novel maths & sciences discoveries and I can run an LLM on my phone for simple tasks that would've been unthinkable a few years ago are testaments to the directionality of the tech.
When this house of cards collapses, AI research dries up, and companies pivot to the next hype cycle there will be a generation of people left with atrophied skills and lingering addiction and psychosis. The most flexible will bounce back just fine, but many will never recover from this damage.
Sincerely, I hope you're in the former category.
Hint: Those are not AI fans. (See the top comment for context.)
interdisciplinary teams/person + SOTA AI + Right context (customer/ internal pain points, access to the code, focussed Hackathon) = amazing cool things.
I’ve had Claude do stuff like interrogate IoT devices, look into proprietary file formats and even assisted debugging electronics from a text description and datasheets. It has stupidly broad knowledge as long as you watch out for things like getting part numbers/boards correct. Have you ever tried to ask for help on an electronics forum full of miserable greybeards? $20/mo for getting ballpark answers immediately is excellent value.
Do assess if you want the learning experience or not, and whether you care if the problem is solved for you.
For writing I get the appeal to bloggers, but I can’t stand reading anything that has the hallmarks of generative AI.
This is 100% true, and this type of building is where AI is adding real value beyond the "I automated my entire business" hype.
Unfortunately, most companies don't have small interdisciplinary teams who have autonomy to scope and ship software. I spent 10 years at IDEO telling clients this was the way, but it's virtually impossible to replicate on the client side.
> Look: I work at an AI company. I use AI all day.
Looks like OP works at dropbox. Dropbox is not an AI company. It's not remotely one.
Why does this matter? Because it undermines the entire point of the post. Later:
> And if you’re watching that kind of hyped content: You can be part of the solution, too. Hold your favorite creators accountable! Ask them to show you the receipts! If you know that something doesn’t work, don’t just let it slide.
AI confidence theater for me, but not for thee.
The challenge in 2026 is finding the company that's not AI-powered (according to themself)
I'm empathetic to their position though. It is entirely unsurprising to me that someone working in a growth role at Dropbox is unimpressed by the current state of AI relative to its broader claims in the market. They're not working on AI itself nor are they using applied AI where you see the biggest gains (e.g. SWE, ML, data science, etc.).
We still have a significant capability overhang at the frontier for a big chunk of knowledge work task domains, so I think its understandable (given the above selection bias) why someone would think the confidence is overblown. They have a point in their own domain.
I do think a small number of people have totally transformed themselves and their business using AI. But that doesn't overlap with the people who are loudest about it on LinkedIn, X, and other channels.
Sure, HN has changed over time, but this tends to be a place where people field technical questions and get technical answers.
It’s marketers doing marketing all the way down.
I have big problems with how AI companies stole artist's art and I have big concerns with how AI might be overused or misused. However, I don't understand how people can still say that AI is useless. Everyday when I use AI I am impressed with everything it can do. Particularly in coding, AI is useful right now (this may or may not be due to how messy the codebase that I'm working in is).
To be clear, this isn't a rhetorical question. I'm wondering what I am missing that other people are seeing.
llms ability to output usable code is insane, I agree. It changed the industry and the shape of what the job means. But the ability to code 10 X more output doesn’t magically create more value. In fact, I am personally skeptical and I’m planning on the first few years of growing pains. These LLMs make the job hard harder, make coordination and collaboration harder. There’s going to be a lot of churn. I quit my last job over the AI mandates.
Also, a lot of people get stuck on the idea that the main use case of LLMs is to generate more production code. I don't see it like that at all anymore. It's just as useful or even more useful as a problem-solving partner, generating throw-away prototypes, conducting experiments, verifying assumptions against code and documentation, performing one-off UI or smoke testing.
“It’s totally useless” to “it hallucinates constantly” to “it’s not revolutionary” to “show me where it’s 100x productivity gains” to “if AI is so good then why isn’t OpenAI one shoting all its software and even cloning all software on the planet?”
For me, the big win is that it's very cheap to experiment with several approaches to something and pick what feels like a good winner. For UX work this is a boon because it shifts the bottleneck to evaluating designs, which is where the bottleneck should be. It has historically not been there.
Did they tell it to write it differently? You can literally make it sound like a pirate if you want. You can also make it be conservative, non-hype. Just ask.
Were it so, I wouldn’t mind so much. But they hurt others, too.
And I agree with the authors thesis that the hype is actively harmful. Specifically (and this is a confession not a judgement) if you’re usage of AI is limited to mindlessly vibe coding tools all day long, your missing the actual process of just fumbling through the awkward stage of being consciously ineffective, so you can break through to eventual productivity. And the real productivity probably isn’t as exciting as the game or app you one shotted for fun.
- I no longer look at log files. If there's an MCP that gets access, I don't even download them
- I don't look up implementation details, just say "I want this bit to do X". In general, I don't concern myself with technical details anymore
- I don't try to debug
The way in which these examples are life changing, is that I now solve abstract problems, not technical ones. Of course I could still do my job if AI disappeared tomorrow, but a road sweeper could also do theirs if the sweeping trucks went away too.
> Of course I could still do my job if AI disappeared tomorrow
So then you agree with TFA, it’s not so critical that if taken away your work would not fall apart.
The main thrust of what he was saying is in the second half: "a road sweeper could also do theirs if the sweeping trucks went away too" - you could do your job tomorrow without an IDE auto-completing common syntax, but it's a useful tool for many people. OP is saying that AI, for him, is similarly a very useful tool that he considers to be close to "critical in [his] day".
Why not actually address that instead of playing word games?
The point is that with this AI technology, we're putting up with a lot of shit -- everything is getting worse across the board -- on the promise of incredible upsides, so where are they? "I'm X% better at my job" is not seriously responsive to "Show me something truly life changing." That kind of technological progress is routine, so why do we have to put up with all the shit for routine technical progress we could get in any number of other ways that don't enshittify everything.
You're the one holding him to a standard that it be "truly life changing" when that was never what he was addressing in his comment. (You're also the one adding emphasis that increases the requirements to overcome the authors, and your own, doubt). And who are you to say that these things didn't truly change the OP comments life? Not having to look up implementation details could be a time save for them on the order of hours per week. That could free up time for him to develop a new hobby, start working out, or any other manner of things that could literally change his life. You're not the only one who gets to decide what life changing means.
Also, it really is worth reflecting on the fact that we are where we are as a society because of a bunch of cumulative "I'm X% better at my job" advancements. That's literally one of the main through-lines in technological advancement since the stone age. Another through-line is that technology (even "truly life changing" technology) is frequently over-hyped. There's really nothing new going on here.
Then maybe he shouldn't have quoted a portion asking for "truly life changing examples that can't be done without AI" with examples that are not lifechanging, and then concede that yes, his job can still be done without AI.
No one argues that AI provides marginal benefits and makes some people's tasks more efficient. But people are noticing that leaders of AI companies have promised everything from "universal high income" to "universal basic compute" and none of that is really happening.
Meanwhile, the big negatives are here now -- job losses, pay cuts, data centers taking over land, poisoning water, and straining electricity grids. If we're going to put up with those downsides, it's reasonable for people to ask where the truly lifechanging upsides are (the whole point of TFA and this entire Hacker News thread), and "I don't have to debug my PRs anymore" is decidedly not that.
>> Of course I could still do my job if AI disappeared tomorrow
> So then you agree with TFA, it’s not so critical that if taken away your work would not fall apart.
That's ungracious, at best. Your interpretation is something close to the death of the position, while theirs is the moderate interpretation of "causes chaos".
Just today I wanted to add to my nvim setup some validation for Kubernetes YAML. Asked Claude why it wasn’t looking at schemas correctly. After 30 minutes of prompts and nudges, during which it went on multiple wild goose chases, I just gave up and did it myself.
Other times I did have more success but it’s not rare to encounter this infinite loop style drill-down, which ultimately wastes more time than if I did it myself from the beginning.
To brag about this openly...
That said, as long as there's still technology, it's awfully nice and a bit gamechanging to have an effective junior engineer who follows orders and who's happy to wade through stack overflow and customize solutions for my problems. It's not that I couldn't do it myself, it's that doing so for one shot weird config issues and arbitrary UI/OS changes sucks the very life blood out of my soul.
AI is a tool, an amazing tool IMO, your own personalized philosophical zombie ready to do your bidding. Social media and mobile devices OTOH, they both suck the life out of me moreso than chasing config issues. Unfortunately, some of the best sources of info about some of my interests are trapped on FB in a swamp of rage bait, sigh.
Marketing has always been about overpromising and underdelivering, no? OK, show me when it wasn't. And if overpromising on hot tickets like the AGI and FSD didn't get people the big bucks and engagement, maybe they'd dial it back a notch, but that's just not how it's played out so far. See the current crop of contenders to disrupt the leading maker of AI HW with their noisemaking CEOs for the evidence thereof.
Marketers exist because every company, every VC, and every stock holder (Wall St. and every person that has money invested for retirement or to pay for their kids college tuition, etc), demands perpetual growth. Till this changes, and it won't any time soon, marketers will exist.
No one is working on changing that. Anyone?
Startups are the worst. No one is hiring a marketer at a startup to drive 25% yearly revenue growth, the expectation is 10X or 100X. How is that supposed to happen without pushing advertising, hard?
We can blame the marketers, but this is structural problem.
This alone will improve the lives of so many people. The real issue with AI commentary is that everyone is guilty of the hedonistic treadmill. We constantly need it to do more and more to get that sense of awe.
Having this hype ingested fully by finance is lethal. Ive witnessed CFOs fully believing a lot of the hype and that directly leads to massive chopping of random teams on a spreadsheet with 'use ai' directives.
My swiss banks won't be able to do this in the next 5 years.
I am not affiliated with wise, just a happy user.
Does the fact that I barely wrote any lines of code in the past six months while my job has been for the past 20 years (and still is) that of producing code, feels "life changing" enough?
It is life changing in the sense that suddenly you realise you did not exist for all those years.
It is not. It can shortcut only because it had been done before.
> Meanwhile, I’m over here asking ChatGPT to rewrite the same paragraph for the third time because it keeps defaulting me into ‘LinkedIn wisdom post’ mode. GARH.
I'm not doubting the author this is their experience, but is this with the trash free/instant ChatGPT or something else? If even "Thinking" was working like that, together with proper prompting, then I'd be surprised by the author's experience.
But until people start showing exact examples and exactly what they're using, all the navel-gazing around this, positive, negative or neutral, will all just be empty words we can't really know what to do with.
I suspect the majority of users won't be aware of what their current setting is.
Even though you're right, we can't know what to do with their words.
Descript. I edit video for work and I am not ever going back to manually seeking to edit video. I am about 8x faster with Descript.
This isn’t even a raw praise post: Descript is wildly unstable and shitty at times but I STILL won’t use anything that doesn’t have transcription-based editing.
(I'm unaffiliated with them.)
The problem with it is that unlike sell-side mortgage fraud, self-driving cars, crypto, and current AI… quantum computing doesn’t have a simple, approachable, mass appeal parlour trick to rally around.
Until it gets one, it’s never going to facilitate the scale of grift and graft as the others.
If all crypto can only be broken by a machine that can itself only be made by 3 companies all of which are very careful about who they sell to, we get a few governments spying on everyone (including each other), while keeping normal criminal hacking (and insufficiently important governments) locked out.
And it’s unrealistic to expect them to stay perfectly motionless, they will try to adapt to the changing landscape too.
In enterprise workflows these are essentially super generalizable NLP and CV models that can be developed and deployed at a fraction of the cost.
This is not what the tech bros promised but it’s still pretty damn good!
Hilarious! People think we are in a socialist economy lol.
No: you don't need the broad masses for any confidence. Put your product to the free market and judge whether people like it. That's how it should work and that's how it always works. No one has to convince Jane Doe that AI enhances productivity by 20% and get her approval before deploying data centers.
But then how will the investors justify having poured over a Trillion USD into AI?
Confidence theater will continue until morale improves.
That's a benign term for it. Actually, it's lies, damn lies, and fucking god-damn lies. It's the super-mega-overdrive version of every principle that ads are based on. Like they are afraid we still haven't understood what makes ads tick, so they give us the practical in-your-face demo. And they were successful. I'm not believing a single word anyone remotely related to that business says anymore.
Louis CK's wifi on a plane bit is funny because it's true.
Oh hell no. A lot of jobs were lost there as well. Marketing got demoted into "just prompt AI" across the board, everyone and their dog built "ingest paper receipts from arbitrary sources into ERP/travel expense programs" (because there still is no standard on "how to transform a paper bill into a QR code"), and HR... "I inserted unreadable white text into resume PDFs to cheat AI resume filters" is reality, not just a meme.
> What I don’t like about it is that it creates a fake baseline. If everyone around you appears to have figured something out about AI that has transformed their work, then using AI to summarize meetings suddenly feels embarrassingly basic.
Meetings themselves and emails have gone the full bullshit circle. AI agents fluff up prompts into Powerpoint slides so incoherent everyone forced to sit through it eventually nods off and reads the AI summary later on. For emails, similarly AI agents fluff up prompts, send it over the wire, only for another AI agent to distill it back into something consumable by a human. Humans creating engaging content by hand seems to be a lost art these days.
Frankly... I can only recommend, leave for some sort of job that is not corporate BS or can otherwise be replaced by a human. Learn a classic trade, to operate heavy machinery or whatever else. Maybe join a firefighter or EMS corps, saving lives is an experience of its own class. Anything IT or corporate is a dead end.
First, you can see at the end of the day with your eyes what you have accomplished (which is way better for your mental health), second, it will take quite some time until there's a robot physically capable of the required dexterity to pull and wire cable and an AI capable enough to coordinate that robot, or find and clean a clog in 100 meters of sewer line. Once AI is good enough to replace a clogged shitter... invest in a good gun and target practice, because the rate things are going, society will break down at that point.
Or, move towards the countryside and raise some chicken, goats or whatever. A ton of tech people have done so in the last years, fed up with the bullshit.
Knowing that you hold gigabits in your hands and only your skill matters is priceless.
But yeah it‘s going to be a skill that will be important for datacenters.
First, we have this section:
But the noise continues on volume 11/10. So this is my desperate plea... Can we PUHLEASE stop this AI Confidence Theater, people? It’s doing more harm than good.
Then immediately after:
This post is sponsored by Firecrawl. Firecrawl is the web data API to search, scrape, and interact with the web at scale. Turn the live web into clean, structured data your agents can actually use.
(P.S. I personally use Firecrawl in my Lovable apps all.the.time - most recently to scrape this very blog so my AI double could use my latest posts as context.)