Use it or lose it, as it were.

0: https://www150.statcan.gc.ca/n1/daily-quotidien/241210/dq241...

This is similar to the photo-taking impairment effect where people will remember an event more poorly if they took photos at the event. Their brain basically subconsciously decides it doesn’t need to remember the event because the camera will remember the event instead.

If the tool is powerful enough to do a better job than our brains would, it's a big win. In fact, we built the entire technological civilization on one such fundamental tool: writing.

Or from another perspective: our brains excel at adapting to the environment we find ourselves in. The tools we build, the technology we create, are parts our environment.

make stock number go up and up and up and people get in the way

I don't think there is that much value in memorizing rarely used, easily looked up information.

Facts alone are like pebbles on a beach, far better (IMO) to have a few stones mortared with understanding to make a building of knowledge. A fanciful metaphor but you know ...

To use another metaphor, you can't REALLY see the forest amongst the trees, if you don't consider the trees themselves.

One of the reasons I like history so much is because, with enough facts accumulated, you can see how one piece of information flows into another - e.g. dates matter, because knowing the precise order in which important events occur helps you determine how those events may or may not have affected each other in the course of their unfolding.

Sure memorizing dates is boring on its own, but putting them in contexts is exciting - you still need to comb the beaches to find the right stones!

I guess an interesting counterpoint to what I said is something like https://en.wikipedia.org/wiki/Phantom_time_conspiracy_theory (and similar) where a grandiose framework tries to fit inconvenient facts into a shape that is entirely invented.

(Can't say time is the limiting factor either -- we're both in HN comments, valuing our own time at zero.)

Let's for example about html boilerplate, where you don't remember the syntax. What you remember is the components & why they are needed, then add them one by one as you recall your memory. Doctype, html tag, head, body, etc. It works because html is simple and common.

Then for express it is harder, because you need to recall javascript syntaxes and express syntaxes, and most of the time you don't get involved with express outside req and res. You recall that express need body parser, register routers, and finally listen, whether you use http server first or directly from express. Now you compose one by one, looking at docs or web for the forgotten pieces, but you don't lose the understanding / logic of express, you just forget the syntaxes.

As for stream where I keep forgetting it, I just need to remember that stream need source, event handler such as on data, error, finish / end. Pipe if needed. However I never remember whether to use writable, readable, streamable, etc because I seldom get involved with them, and can look up for references anytime.

There is limited time, of course - no one can learn everything, but you can pay attention to the important facts, and the connections between them.

In some ideal world you would learn every fact there is, and the connections would fall out on their own, but in the real world we have to construct theories and frameworks to organise facts.

And of course, what if your phone dies?

2 years ago you had downloaded onto your laptop an effective and useful summary of all of the information on the Internet, that could be used to generate computer programs in an arbitrarily selected programming language.

I wrote a post about it: Your toaster will know mesopotamian history because it’s more expensive not too.

https://wanderingstan.com/2026-03-01/your-toaster-will-know-...

But will it know the difference between too and to?

And as other commenter pointed out, a smart toaster with ads or data collection can be subsidized and thus be more profitable. (Oh what a world we're headed for!)

In any case, I think the LLM-everywhere thesis holds even strong for even moderate-complexity devices like power plugs, microwaves, and mobile phones.

I'm sure people would get a cheaper toaster in exchange of an ad being burned in your bread.

Ask your local model a verifiable question - for example a list of tallest buildings in Europe. I did it with Gemma on my laptop, and after the top 3 they were all fake. I just tried that again with Gemma-4 on my iphone, and it did even worse - the 3 tallest buildings in Europe are apparently the Burj Khalifa, the Torre Glories and the Shanghai Tower.

I wouldn't call that effective compression of information.

As another commenter said, the affordability of LLM subscriptions (or, as others are predicting, the lack thereof) is the primary concern, not the technology itself stealing away your skills.

I am far from the definitive voice in the does-AI-use-corrupt-your-thinking conversation, and I don't want to be. I don't want LLMs to replace my thinking as much as the next person, but I also don't want to shun anything useful that can be gained from these tools.

All that said, I do feel that perhaps "dumber" LLMs that work on-device first will allow us to get further and be better, more reliable tools overall.

Were people actually physically typing every character of the software they were writing before a couple of years ago?

I on the other hand am a software engineer, so writing code is part of the job title.

Company? Helm? Whatever vi uses that's like company and helm? Haven't IDEs written function calls for you for like decades now?

I work on greenfield projects so I see my fair share of boilerplate, but honestly, it's just a minute part of work that's almost meditative to write a little bit of trivial code (i.e. a function signature) in between sessions of hard thinking. Writing boilerplate is very far down the list of things I seek to optimize.

Also I don't use LLM.

I'm not super worried, either I still do the last leg of the work, or I go back an abstraction level with my prompts and work there

I remember at that time, my "mentor" suggested to memorize all the "keywords" from C (which were few). But given my bad memory I had to constantly look at the book.

Aaah how times have changed.

[1] https://openlibrary.org/books/OL1601323M/Turbo_C_C

I learned the Q array language five years ago and then didn't touch it for six months. I was surprised how little I remembered when I tried to resume.

> I cannot remember basic boilerplate stuff.

I don't know exactly what you mean by boilerplate stuff, but honestly, that's stuff we should have automated away prior to AI. We should not be writing boilerplate.

I'd highly encourage you to take the time to automate this stuff away. Not even with AI, but with scripts you can run to automate boilerplate generation. (Assuming you can't move it to a library/framework).

Last time I looked there were at least seven ways to do it.

I'm wondering if this is something that hits new developers faster than more experienced ones?

Played with these coding agents for the last couple weeks and instantly noticed the brainrot when I was staring at an empty vim screen trying to type a skeleton helloworld in C.

Luckily the right idioms came back after couple of hours, but the experience gave me a big scare.

Very interesting, wonder what makes our experiences so different? For you "playing for a couple of weeks" have a stronger effect than for me after using them almost exclusively for more than a year, and I don't think I'm an especially great programmer or anything, typical for my experience I think.

Maybe the difference between actually knowing stuff vs surface level? I know a lot of devs just know how to glue stuff together, not really how to make anything, so I'd imagine those devs lose their skills much faster.

I do still write stuff manually frequently - I often spend 5 minutes writing structs or function signatures to make sure that the LLM won't misunderstand or make something up. Maybe that's why I haven't lost it.

Almost certainly, at least according to Ebbinghaus' forgetting curve.

Right, sounds very credible to me. What did you write, an addition function in each of those?

I wrote a little toy-calculator in each, ended up being ~250 LOC in each of them, not exactly the biggest test but large enough to see if my muscle memory still works which I was happy to discover it still did.

> But now, I'm useless. My mind has turned to pudding. I cannot remember basic boilerplate stuff

> Played with these coding agents for the last couple weeks and instantly noticed the brainrot when I was staring at an empty vim screen trying to type a skeleton helloworld in C.

This is very different from what I'm (not) experiencing. My test was for if I can remember the basic syntax of the language itself, I was never a big framework user, so of course using a framework is about the least interesting test I could do of myself.

Instead, I did the bare minimum to see if my "mind has turned to pudding" or "instantly noticed the brainrot", which would have been visible even for a toy calculator, obviously.

> the people pushing for "AI writing all the code" are not worried about you retaining the "muscle memory" to write addition and substraction functions

What are they worried about then? From your perspective, sounds like they're worried about "using these and those frameworks" but that's far from "real world work" in my experience, and really the least interesting thing you could remember as a developer.

People using AI for tasks (essay writing in the MIT study linked below) showed lower ownership, brain connectivity, and ability to quote their work accurately.

> https://arxiv.org/abs/2506.08872

There was a MSFT and Carnegie Mellon study that saw a link between AI use, confidence in ones skills, confidence in AI, and critical thinking. The takeaway for me is that people are getting into “AI take the wheel” scenarios when using GenAI and not thinking about the task. This affects people novices more than experts.

If you managed to do critical thinking, and had relegated sufficient code to muscle memory, perhaps you aren’t as impacted.

My theory is that if you're not full-time coding, it's hard harder to remember the boiler plate and obligatory code entailed by different SDKs for different modules. That's where the documentation reading time goes, and what slows down debugging. That's where agent assisted coding helps me the most.

As an example I have been fighting with agents re-writing or removing guard clauses and structs when dealing with Mach-o fat archives this week, I finally had to break the parsing out into an external module and completely remove the ability for them to see anything inside that code.

I get the convenience for prototyping and throwaway code, but the problem is when you don’t have enough experience with the quirks to know something is wrong.

It will be code debt if one doesn’t understand the core domain. That is the problem with the confidence and surface level competence of these models that we need to develop methods for controlling.

Writing code is rarely the problem with programming in general, correctness and domain needs are the hard parts.

I hope we find a balance between gaining value from these tools while not just producing a pile of fragile abandonware

i dont think it is entirely a case of voluntary outsourcing of critical thinking. I think it's a problem of 1) total time devoted to the task decreasing, and 2) it's like trying to teach yourself puzzle solving skills when the puzzles are all solved for you quickly. You can stare at the answer and try to think about how you would have arrived at it, and maybe you convince yourself of it, but it should be relatively common sense that the learning value of a puzzle becomes obsolete if you are given the answer.

The people who used chatGPT had the most difficulty quoting their own work. So not boilerplate, CRUD - but yes the advantage is clear for those types of tasks.

There were definite time and cognitive effort savings. I think they measured time saved, and it was ~60% time saved, and a ~32% reduction in cognitive effort.

So its pretty clear, people are going to use this all over the place.

I do use AI daily to help me enhance code but then... I also very regularly turn off, physically, the link between a sub-LAN at home and the Internet and I still can work. It's incredibly relaxing to work on code without being connected 24/7. Other machines (like kid's Nintendo switch) can still access the Internet: but my machines for developing are cut off. And as I've got a little infra at home (Proxmox / VMs), I have quite a few services without needing to be connected to the entire world: for example I've got a pastebin, a Git server, a backuping procedure, all 100% functional without needing to be connected to the net (well 99% for the backuping procedure as the encrypted backup files won't be synch'ed with remote servers until the connection is operational).

Sure it's not a "laptop on a plane", but it's also not "24/7 dependent on Sam Altman or Anthropic".

I'll probably enhance my setup at some point with a local Gemma model too.

And all this is not mutually exclusive with my Anthropic subscription: at the flick of a switch (which is right in front of me), my sub-LAN can access the Internet again.

The fact that this is being called out is strange.

So I don't think this is all AI tbh.

Repeat this for decades, and "very basic programming tasks" might be creating a cross-platform browser by using LLMs via voice dictation.

Yes, you lost some abilities. Install local model so you have someone to talk to while you are on the plane ;)

No we don't and we never should actually, compilers need to be deterministic.

Also, give the same programming task to 2 devs and you end up with 2 different solutions. Heck, have the same dev do the same thing twice and you will have 2 different ones.

Determinism seems like this big gotcha, but in it self, is it really?

"Do the same thing" I need to be pedantic here because if they do the same thing, the exact same solution will be produced.

The compiler needs to guarantee that across multiple systems. How would QA know they're testing the version that is staged to be pushed to prod if you can't guarantee it's the same ?

With the right settings, a LLM is deterministic. But even then, small variations in input can cause very unforeseen changes in output, sometimes drastic, sometimes minor. Knowing that I'm likely misusing the vocabulary, I would go with saying that this counts as the output being chaotic so we need compilers to be non-chaotic (and deterministic, I think you might be able to have something that is non-deterministic and non-chaotic). I'm not sure that a non-chaotic LLM could ever exist.

(Thinking on it a bit more, there are some esoteric languages that might be chaotic, so this might be more difficult to pin down than I thought.)

No one trusts Congress or the US government to effectively regulate AI for the greater good of the population. Each party believes regulations proposed by the other party will be used to discriminate against and control their party.

But online (because the available search criteria are so imprecise) your search brings up every possible form of clothing, especially the stuff that's a commodity (or hyped by major e-merchants) -- cheap, popular with 25 year olds, colorless, largely disposable. It's hopeless unless you yourself are a commodity -- indiscriminating, predictable, and totally average.

> I couldn't imagine going through items video by video.

That's fair, it's just not how people use it and it's not the concept. It's primarily a browse experience, not a search experience. You can search but that's not the core experience.

I buy vinyl records and retro games. There are sellers that I like. When I open the app I see which of my preferred sellers are live and I tune into their stream and hang out and watch them. If something I'm interested in pops up, I'll bid on it. Live shopping is not trying to be "ebay but video", it's a different experience.

https://en.wikipedia.org/wiki/HSN

Could be an advantage to knowing this because you can step around it.

_You_ know it's AI, so you go in person to a store. Likewise, next time you hire, you can simply refuse to accept "cover letters".

I was already getting disillusioned with the Internet as a learning resource during the SEO spam era, but the AI era has completely destroyed it.

Education and targeted summary searches are one of the best uses. I literally found the location of the criminal who embezzled thousands of euros from my condominium with an AI search. It took me around fifteen minutes. Other people had been looking for years. (True story...)

Obviously none of that is happening in the current moment, and I grant that cake recipes would be low stakes, but I would rather take the tradeoff of trawling through a little bit of slop to get that same information than acclimate myself to a workflow that could be abused by providers in more high-stakes situations down the line.

But that's just me, and I realise this is not a particularly popular take, but it should nonetheless be illustrative for why "just ask the LLM" might not be the best of ideas long term.

The article's US-China comparison quietly reveals the prediction that would follow from the thesis: if the Perez 'late deployment' framing is right, then the Chinese model—lean, industrial, healthcare and education application, grounded in near-term ROI—is betting correctly on where we are in the curve and should outperform over the next decade. That's a concrete, testable claim that would validate or falsify the argument independently of whether AI constitutes a 'new surge.'

There are some very interesting information network theories that present information growth as a continually evolving and expanding graph, something like a virus inherent to the universe’s structure, as a natural counterpoint to entropy. And in that view, atomic bonds and cells and towns and railroads and network connections and model weights are all the same sort of thing, the same phenomenon, manifesting in different substrates at different levels of the shared graph.

To me, that’s a much better and deeper explanation that connects the dots, and offers more predictive power about what’s next.

Highly recommend the book Why Information Grows to anyone whose interest is piqued by this.

1) as in the article it's a contraction of work- industrialization getting rid of hand-made work or the contraction of all things horse-related when the internal combustion engine came around

but- it will also be

2) new technologies and ideas enabled by a completely new set of capabilities

The real question is if the economic boost from the latter outpaces the losses of the former. History says these transitions aren't easy on society.

But also, the AI pessimism is hard to understand in this context- do people really believe no novel things will be unlocked with this tech? That it's all about cost-cutting?

Many people claim its going to become a tool we use alongside our daily work, but its clear to me thats not how anybody managing a company sees it, and even these AI labs that previously tried to emphasize how much its going to augment existing workforces are pushing being able to do more with less.

Most companies are holding onto their workforce only begrudgingly while the tools advance and they still need humans for "something", not because they're doing us some sort of favor.

The way I see it unless you have specialized knowledge, you are at risk of replacement within the next few years.

The problem is that there are not many fields that are going to be immune to AI based cost cutting and there surely will not be enough work for all of us even if we all retrain.

If we all do, then it will create a n absolutely massive downward pressure on wages due to massive oversupply in other lines of work too

So there's really just no good way out

Well, it really isn’t. First, this entire post makes two assumptions: 1) that AI adds more value to the process than it removes and 2) that it’s sustainable.

It’s not pessimism to want to validate these first.

Are AI “gains” really transformative or simply random opportunities for automation which we can achieve by other means anyway?

Can the world continue to afford “AI as a service” long enough for the gains to result in improvements that make it sustainable? Are we dooming our kids to a hellishly warm planet with no clear plan how to fix it?

It’s not pessimism, just simple project management if you ask me.

They're transformative in the sense that will shrink the optimal team size, but I don't expect the jobs to actually go away unless these things both get substantially better at engineering (they're good at generating code but that is like 20% of engineering at best) and we have a means of giving them full business/human levels of context.

Really basic stuff gets a lot easier but the needle doesn't move much on the harder stuff. Without some sort of "memory" or continuous feedback system, these models don't learn from mistakes or successes which means humans have to be the cost function.

Maybe it's just because I'm burnt out or have a miner RSI at the moment, but it definitely saves me a bit of time as long as I don't generate a huge pile and actually read (almost) everything the models generate. The newer models are good at following instructions and pattern matching on needs if you can stub things out and/or write down specs to define what needs to happen. I'd say my hit rate is maybe 70%

Trust me, this is a work in progress. Right now most corporations do not have their data organized and structured well enough for this to be possible, but there is a lot of heat and money in this space.

Imo, What most of the people that are not directly working in this space get wrong is assuming swes are going to be hit the hardest: There are some efficiency gains to be won here, but a full replace is not viable outside of AGI scenarios. I would actually bet on a demand increase (even if the job might change fundamentally). Custom domain made software is cheaper as it has ever been and there is a gigantic untapped market here.

Low complexity to medium complexity white colar jobs are done for in the next decade through. This is what is happening right now in finance: if models stopped improving now, the technology at this point is already good enough to lower operational costs to the point where some part of the workforce is redundant.

I think you misunderstand what I'm saying. I'm not really referring to data systems at all, I'm referring to context on what problems are actually being solved by a business. LLMs very clearly do not model outcomes that don't have well-defined textual representations.

I'm not sure that I agree with white collar jobs being done for, not every process has as little consequence to getting it wrong as (most) software does.

Yeah i misunderstood your point, i completely agree with what you are saying.

I honestly do not believe that strategy, decision making and other real life context dependent are going to be replaceable soon (and if it does, its something other than llms).

> I'm not sure that I agree with white collar jobs being done for, not every process has as little consequence to getting it wrong as (most) software does.

Maybe im too biased due to working in a particularly inefficient domain, but you would be surprised how much work can be automated in your average back office.

Much of the operational work is following set process and anything out of that is going to up the governance chain for approval from some decision maker.

LLM based solutions actually makes less errors than humans and adhere to the process better in many scenarios, requiring just an ok/deny from some human supervisor.

By delegating just the decision process to the operator, you need way less actual humans doing the job. Since operations workload is usually a function of other areas, efficiency gains result in layoffs.

> Much of the operational work is following set process and anything out of that is going to up the governance chain for approval from some decision maker.

Oh that's very interesting! Thank you for the insights!

This is exactly what people were saying a decade ago when everyone wanted data scientists, and I bet it's been said many times before in many different contexts.

Most corporations still haven't organised and structured their data well enough, despite oceans of money being poured into it.

If they've shrunk the team size, that means some jobs (in terms of people working on a problem) will have gone away. The question is, will it then make it cheap enough to work on more problems that are ignored today, or are we already at peak problem set for that kind of work?

Spreadsheets and accounting software made it possible to have fewer people do the same amount of work but it ended up increasing the demand of accountants overall. Will the same kind of thing happen with LLM-assisted workloads, assuming they pan out as much as people think?

If if people actually listened to the people wailing "but what about the horse carriage business!!!" in the 20th century, it would have been a disaster.

it would be like an oncologist trying to stop an anti-cancer pill because theyd be out of a job.

a) just hand wave away that there is a massive power and wealth differential involved in this "mutual agreement" b) dismiss all discussions which recognize that fact... as "outdated Marxism"

Plenty of mainstream economists are capable of seeing the real world which you are pretending doesn't exist.

Even Marx meant the word "exploit" in relatively value neutral terms, just recognizing that in any economy built on private property we exploit humans the same way we do any "resource". It's up to the reader whether they see that as having any moral connotation.

>Plenty of mainstream economists are capable of seeing the real world which you are pretending doesn't exist.

Not really. There are total of zero economic policies made by analysing economy through Labour Theory of Value or whatever other crap Marxists believe.

The above poster used "exploit" in non value neutral terms. Marx tried very hard to be value neutral about it (but its clear what the intentions were) but his readers don't play that game.

The cost cutting is the only revenue producing models for the AI companies so far. It's being pitched as a way for corporations to fire a lot of employees and save money.

Revenue for the consumer facing products is not very impressive. Consumers are mostly satisfied with the free versions and very resistant to adding yet another channel to shove advertising at them.

If you step back a little, a lot of people simply don't see the forest for the trees and they start imagining bad outcomes and then panic over those. Understandable but not that productive.

If you look at past changes where that was the case you can see some patterns. People project both utopian and dystopian views and there's a certain amount of hysteria and hype around both views. But neither of those usually play out as people hope/predict. The inability to look beyond the status quo and redefine the future in terms of it is very common. It's the whole cars vs. faster horses thing. I call this an imagination deficit. It usually sorts itself out over time as people find out different ways to adjust and the rest of society just adjusts itself around that. Usually this also involves stuff few people predicted. But until that happens, there's uncertainty, chaos, and also opportunity.

With AI, there's going to be a need for some adjustment. Whether people like it or not, a lot of what we do will likely end up being quite easy to automate. And that raises the question what we'll do instead.

Of course, the flip side of automating stuff is that it lowers the value of that stuff. That actually moderates the rollout of this stuff and has diminishing returns. We'll automate all the easy and expensive stuff first. And that will keep us busy for a while. Ultimately we'll pay less for this stuff and do more of it. But that just means we start looking for more valuable stuff to do and buy. We'll effectively move the goal posts and raise the ambition. That's where the economical growth will come from.

This adjustment process is obviously going to be painful for some people. But the good news is that it won't happen overnight. We'll have time to learn new things and figure out what we can do that is actually valuable to others. Most things don't happen at the speed the most optimistic person wants things to happen. Just looking at inference cost and energy, there are some real constraints on what we can do at scale short term. And energy cost just went up by quite a lot. Lots of new challenges where AI isn't the easy answer just yet.

At some point those became almost fully obsolete in a productive economical sense (they're just fancy toys now, basically). No 'raising the ambition' is ever going to change that. They are what they are and they can do what they can do.

I don't know about you, but if the something in "we'll find something to do" is becoming a toy for AI or very rich people, I'm not exactly hopeful about the future.

As for rich people, they get richer and richer until people correct them. Sometimes violently. The current concentration of wealth in particularly the US seems more related to political changes since about the Reagan era than to any recent innovations related to technology.

This is false. Being fatalistic and 'panicking' can definitely influence and thus change the outcome. Your logic is similar to what is (incorrectly) used to dismiss the Y2K-problem, for instance: Looking back it seems like there was no need to panic, but that is only because a lot of people recognized the urgency, worked their ass off and succeeded in preventing shit from going horribly wrong.

See: https://en.wikipedia.org/wiki/Preparedness_paradox

Your handwaving is doing harm by lulling people into a false sense of security. Your initial comment amounts to "Ah, it'll be fine, don't worry about it. We'll adapt, we always have.", even though you provide absolutely no arguments specific to this enormous force of insanely rapid change in an already incredibly unstable fragile world. We might adapt, but it will require serious thought rather than handwaving and leaning back; even then it might come with massive societal upheaval and a lot of suffering.

A lot of people seem to be wasting a lot of energy insisting it is all going to end in tears because <fill in reasons>. All I'm doing here is pointing out that people like this come out of the woodwork with pretty much every big change in society and then people adapt and things are society fails to collapse.

I'm not arguing there won't be changes and that they won't be disruptive to some people. Because they will and people will need to adjust. But I am arguing that a lot of the dystopian outcomes are as unlikely to happen with this particular change as they have been with previous rounds of changes. I just don't see a basis for it. I do see a lot of people who want this to be true mainly because they are afraid of having to adapt.

> already incredibly unstable fragile world

There are a lot of people arguing that things are better than ever by most metrics you might want to apply for that. The reason you might feel stressed about the news is that dystopian headlines sell better and you are being influenced by those. That's also why the Y2K got a lot more attention than it deserved in the media and then a lot of people indeed freaked out over that. Of course a lot of that got caught up in people believing for other reasons we are all doomed and that the apocalypse was coming. And it made for amusing headlines. So, it got a lot more attention than it deserved. And then the clock ticked over and society failed to collapse.

Yes. It's a mostly shitty but very fast and relatively inexpensive replacement for things that already exist.

Give your best example of something that is novel, ie isn't just replacing existing processes at scale.

It's been 3 and a half years now since the initial hype wave. Maybe I genuinely missed the novel trillion dollar use case that isn't just labor disruption.

There are a lot of really useful things that were impossible before. But none of these use cases are "easy," and they all take years of engineering to implement. So, all we see right now are trashy, vibe-code style "startups" rather than the actual useful stuff that will come over the years from experienced architects and engineers who can properly utilize this technology to build real products.

I'm someone who feels very frustrated with most of the chatter around AI - especially the CEOs desperate to devalue human labor and replace it - but I am personally building something utilizing AI that would have been impossible without it. But yeah, it's no walk in the park, and I've been working on it for three years and will likely be working on it for another year before it's remotely ready for the public.

When I started, the inference was too slow, the costs were too high, and the thinking-power was too poor to actually pull it off. I just hypothesized that it would all be ready by the time I launch the product. Which it finally is, as of a few months ago.

It's kind of like dealing with Amazon, or any other company that has both compute and the ability to sell the kind of product you make.

Said AI providers can sell you the compute to make the product, or they can make the product themselves with discounted compute and eat all the profits you'd make.

However, my product is so far removed from anything these companies would make, on top of that I'm using open-source models (e.g., oss gpt 120b is really, really good). I don't use any of the main providers like AWS, etc., and the underlying AI systems are only about 5% of the product. I need it for the idea to work, but it is a tiny part of the full offering. I can't really imagine it would make any sense for Amazon, etc., to compete on something like this.

But yes, in the end, huge conglomerates with infinite money can destroy smaller entrepreneurs - but that's not really any different than it's been for decades pre-AI.

I don't really get people who are dismissive about this aspect of AI- my original question wasn't about cost-efficiency of developing these things, but just that the technology itself is creating things that wouldn't have been possible before. It seems hard to refute.

Whether or not it's worth the cost is a different debate entirely- about how tech trees are developed and what the second order effects of technology are. There are so many examples- the computer itself, nuclear power, etc. I think AI is probably on the same order as these.

The implication of your comment seemed to be that this was going to be so much more than replacing people. But I fail to see how any of the items you listed are anything other than that.

These things have always been possible. Just slow and limited by labor. Which is the primary and novel "unlock" of AI.

You can argue it's a good thing, and in many areas I'd probably agree. I'm directly responding to your skepticism and implied absurdity that replacement is the main unlock here. It absolutely is.

Yes, you are off-base.

Solutions to the protein folding problem existed before, but not in the way you are implying.

I do still believe the main value proposition is large scale replacement and am unconvinced that most people driving AI adoption have these other more noble pursuits in mind with respect to AI.

But I will absolutely stand corrected here and if our dystopian future includes some genuinely useful medicinal advancements then maybe that will make the medicine (heh) go down easier.

But also you don’t need SOTA frontier models for that!

Wouldn't that apply to most technological advances? Cars, computers, cell phones.

e: Also I don't know that I'd strictly bucket these specific examples you gave as shittier versions, though I guess that's a matter of perspective.

One reason people forget that "good quality" shoes existed was that you could only afford to buy one pair ever, not that things were made better, necessarily. (or could be both, but that replacing a pair of shoes was a financial hardship, because hand-made things, even back then, were expensive).

Even if you're against fast fasion I don't think anyone wants a pair of shoes to cost $10,000.

It's always win some, lose some with the economy, but technology itself opens previously impossible capabilities.

Your comment got me thinking about if technology is actually better, but that's a whole new discussion. We wouldn't need the fireproof furniture if we all used the local sweat lodge for bathing or the mess hall yurt for cooking. We wouldn't need wind-resistant clothing if we didn't make personal rockets that go 200mph to travel long distances to arrive at the same amenities (just in a different city).

I'd generally agree, but there are always caveats. See e.g. glass vs. plastic bottles - glass looks like strictly superior solution environmentally, until you consider how much fuel is saved across entire logistics chain by plastic bottles being significantly lighter.

> We wouldn't need wind-resistant clothing if we didn't make personal rockets that go 200mph to travel long distances to arrive at the same amenities (just in a different city)

FWIW, I was thinking more about people who like to walk around in windy places, including mountains, etc. But even if we exclude tourists, we're still left with people who work at altitudes (including infrastructure anywhere - get on a high enough pole or roof, it's going to be windy). More generally, there are people doing useful work, including construction, services, and research, in all kinds of extreme environments, and this is directly enabled by post-industrial era fabrics.

It's essentially a political energy source. It heats everything up.

Eventually it either explodes, goes through a phase change to a new (meta)stable state, or collapses back to a previous state.

Better autocomplete/autocorrect, "circle on screen" -> OCR anywhere, high-quality automated background removal[0].

> fireproof materials

No big examples to point at now, except maybe whatever security fixes that'll come out of Glasswing Project[1].

> So far the only product AI is producing is layoffs.

AI-related[2] layoffs are a direct consequence of useful things AI is delivering.

--

[0] - Super useful for e.g. making ID photos, which I notice I need to do increasingly often, which is likely a consequence of proliferation of remote/digital ID verification, which nicely ties us back to question 'butlike expressed, i.e. how much is technological progress actually improving things.

[1] - https://www.anthropic.com/glasswing

[2] - As opposed to "we wanted to lay off people anyway, AI just provided a socially-acceptable excuse".

From a consumer perspective, AI isn't really producing any new products with real market demand. Chatbots are fun, but there's no indication consumers are willing to pay for them.

I frankly do not care how much novel stuff is "unlocked" with AI tech if it means I become unemployable due to it replacing all of my skills

This is a burden of proof inversion: historically new technology has not resulted in optimistic outcomes. Quality of life improvements were side effects of capital accruing. AI optimism is the naïve option that requires justification.

Cost cutting has less uncertainty than making something new, so they do that first. If something else comes along, then great.

This is also why the people should make the transition as difficult as possible for companies doing layoffs when the companies are paying proportionally very little in taxes compared to the people they are laying off.

AI so far has really only shown massive utility for programming. It has broad potential across almost all knowledge work, but it’s unclear how much of that can be fulfilled in practice. There are huge technical, UX and social hurdles. Integrating middle brow chatbots everywhere is not the end game.

I don't know if this is the effect of relying on AI too much in my day-to-day work or leading a more monotonous life as of late, but I'm sure I'm not the only one. Lots of ideas that I could have built before LLMs took over now seem trivial to build with Claude & friends.

I don't think it's intrinsically wrong, we are in a late stage of a transformation. Software is eating the world and AI is (so far) most profitably an automation of software.

There is plenty of money to be made along the way. I don't really buy the article's seeming confusion about where the money is going to come from. Anthropic is making billions and signing up prodigious amounts of recurring revenue every month.

True but in common parlance "AI" has come to mean LLMs.

> Software is eating the world

The premise of the article is software ATE the world and there isn't much left to eat that hasn't been eaten.

>> At the early stage of a surge, investment tends to be patchy and not fully understood—the sector exists but it is not completely legible yet.

He says this in the context that AI clearly doesn’t fit this pattern, as the investment has been enormous.

I feel like he and everyone else has a scale problem, due to the tendency to equate AI to LLMs - the investment is patchy and not fully understood - I really don’t think we’ve seen anything more than the pretremors at this point - as the scale of the change is just as incomprehensible to the world at large now as it was when the steam engine was just a slightly better way of getting water out of a mine than a donkey.

All hypothetical, but if compute + AI research continues at pace, in 5 years we should see extremely good local models.

Because the first situation won't create that many jobs. The second one might.

The human form factor - torso up anyway - is probably easier to bootstrap on a general basis; keyed off of human data. But I don't like the failure modes of bipedal robots - imagine a robot flailing around trying to regain balance, in any setting with humans around.

I'm no expert of course, just pontificating.

They will keep bleeding money by the way.

In the end, it'll probably require something like model-based RL like Yann LeCun talks about and that's totally different to the LLMs.

Sadly, I think there's a risk we might also be heading towards a dark age with few advances since fundamental research has been squeezed away for being unprofitable or hobbled by a industrialized publishing/review-system for a while now and we've been coasting along on profitable applications rather than (expensive) breakthroughts in basics.

Robotics? lights-out operations in automated factories are already a thing, so I don't know if they're the "next thing".

mRNA vaccines? Sure, they're a huge medical advance. With great potential, in that area. But it's just an area.

Space? Maybe, if we get past LEO, find something useful to do there, and don't succumb to Kessler syndrome.

Eh, I do think this is kind of underestimating the changes in robotics that are occurring. LLMs incorporated with other ML kernels extend the capabilities a long way. That and the amount of computing power now usable to train robotics is far far larger.

I don't see the immediate application of language generation to navigating and manipulation in the physical world.

Show HN: RedSOC – adversarial evaluation framework for LLM-integrated SOCs

Honest question, I'm not trying to mock economists or anything like that.

My wife is absolutely not technical, and she began using ChatGPT before me.

This is to say, I believe you to be correct here. The LLM adoption rate is many times the computer adoption rate. Non-technical people are immediately seeing the benefit of LLMs where they did not with computers in the 1970s.

it costs you nothing to install/adopt an AI chat bot and it's being force fed to everyone at head turning loss in order to justify the push.

If 3D printers could’ve given usage away for years directly in our homes then I bet we would’ve seen wider adoption there too.

People aren’t going to wait minutes per response for clearly inferior results compared to what they get for free on ChatGPT in browser in seconds, whether it’s logical or not. Not to mention they can’t ask more than a few questions tops before the whole thing crumbles. Expectations and reality are too far apart here.

Let’s also address another real issue: what are they going to use? LM studio? Is that really a user experience most will tolerate?

Separately, I have a local camera repair shop and my friend told me its 2 months backlog to get your film based camera worked on.

Ultimately if the deal we get online is infinite tracking, infinite scrolling and infinite enshittification, real life start to sound a whole lot better.

I gladly pay the (modest/token) late fees to help keep them open at this point. If someone set up a local arcade man…I’d be in heaven ha