The tech sector is the most software-defined and virtual of all sectors, it's natural that they are first to change. Other industries need to make capital investments to adapt to change.
You can see a similar history with electrification. A few industries went quickly, but then it took like a decade of little growth before the true benefits showed up. As an example, take a factory that was built around hyrdo-power like a grain mill. You can't just take an electric motor and drop it in in place of the water wheel. It doesn't fit, the wheel is outside, there isn't room between the mill and the wall, etc. Only when you completely rebuild a new mill building designed around the features and limitations of the electric motor do you see the major benefits.
Same thing here; it will be new factories and new companies that are designed from the ground up around AI and its benefits and limitations that really move the needle.
Cool analogy, but there's another possibility: software is uniquely suited for being boosted by LLM-based coding agents. That's not to say LLMs are useless in other domains, but they may not yield as much benefit.
Software creation tools have been worked on for decades to keep fallible human coders out of trouble as much as possible. Starting with compilers that enforce or at least encourage coding style humans can understand. Compilers embody formal definitions of language syntax. Lint and other static analysis technologies contribute by identifying problematic patterns in code. Leak detectors identify memory management problems. Profilers identify performance problems.
Few other endeavors have this much tooling that serendipitously happens to help help LLM-based agents be more useful and less error prone.
A lot of discussion around AI is very high-level and abstract. It helps to dig in and identify the characteristics that actually make a domain more or less amenable to being boosted by AI.
Yeah, I think it's at the level of self-driving Excel.
Now, Excel also makes mistakes while trying to get clever - remember the biologists who had to rename several genes to prevent Excel treating them as dates - but it's also a ubiquitous business tool. And you can combine them. But equally Excel hasn't replaced every white collar job.
I don't doubt it. Earlier today I got an estimate for the revenue, profit margin, and headcount for a potential supplier of a hardware component to a product I'm working on. It would've taken a couple hours of really tedious work for me to do that with Google translate, a search engine, and various market research sites. The result is I know more what I can expect from that vendor in terms of customization and support.
But those are the kinds of estimates that I can eyeball and be more than 80% certain they are plausible. That won't cut it in a lot of domains.
Yeah but theres also no null test here. He might have done anything and still been successful. He might have done something else more optimal and done even better. We can't really tell from a single anecdote.
Too true. Even on the software side i see people blindly “claude make it faster with no mistakes” and gushing about it when the output is genuinely not good not because claude isnt good but because the user of it isnt.
We got a new a CNC machine and everyone is carving away marveling at how easily it chips away at metal when the result is a jagged mess.
> Cool analogy, but there's another possibility: software is uniquely suited for being boosted by LLM-based coding agents.
Or the pretty obvious third possibility: The tech sector is the part selling the LLMs. They're not digging up more gold with their cool shovel technology, they are selling the shovels.
And whole repositories with commit history that provides step by step training on how to code.
AI could probably write decently if every scraped book had all the revisions to create it embedded within or art had every brush stroke documented. As it stands programming is the first to be trained. The rest will take some more time.
I can't speak for everywhere else, but here in the Bay Area (California), pretty much all hospitals use AI for transcription, summarizing, and some other stuff that I'm not fully aware of (I should really listen to what my Nurses say before agreeing to being recorded).
My doctors office has started doing that and it throws the wildest crap into my chart. Diagnosis the doctor didn't mention, personal anecdotes that doc told me about their family become my family medical history. Test supposedly done that weren't.
Ive been trying to figure out some persistent si joint pain and my chart is full of me passing or failed tests nobody performed, observations nobody made and diagnosis that no doctor confirmed.
Its just so much wrongness, ive tried getting it corrected a few times when it really matters but its just so much wrong crap
It is most impactful in tech because tech is almost nothing but languages and more so super expensive languages where one hour of an “authors” time can cost thousands with a supply demand issue so bad workers are imported from across the world.
Artists were already starving. The more physical trades are…physical not language so a plumber still plums. Maybe now they ask gemini instead of google tho.
Areas related most to…language are the most impact by large language models? The one that i think is just as impacted is education because it is also almost all language driven.
I see zero sign of any of the new mills coming around. Steel, crops, farming, none of it has changed and what little has is from…the software side or…language.
The SV world really over indexes on how much software can do.
There are some things that current generation LLMs are extremely good at. They also suck for a lot of reasons and appear to be prohibitively expensive to train as "general" models.
hey are amazing tools, but every time I see a task that LLMs helped me complete in 30% less time, I see 5 more tasks that took me 30-50% longer to complete because of errors, lies, mistakes, hallucinations, context limits.
Good reference. It is going to slowly change things in 20 years we'll look back and a lot of things will be wildly different because of AI systems. But its rate limited by institutions and momentum and human psychology and the way things change will be somewhat unpredictable with unintended negative and positive consequences.
Yeah, it is not going to be slow. Progress compounds, everything is being built on top of fast computers and fast widespread internet connection. You are not at the beginning of the curve like you were in the 20th century where people didn't even know how to use a keyboard and we are already seeing widespread disruption in security.
Literally the entire premise of the article is that it's going slow everywhere but in software. To think otherwise is to be really myopic about how the rest of the world outside of tech operates. And in any case 20 years isn't that slow. The only things that happen faster that that will be things like adobe goes out of business. And let's say AI helps us develop medicine and robotics and better factories and all the other things that would have a big impact a lot faster.. just human trials could take a decade before you really see that much change. Will it have a wider impact faster than computing its self, or the internet? Maybe? Maybe not.
If a lot of jobs were still manufacturing jobs I would agree, but we are in a service economy that uses computers as the main tool. So I don't see how it won't come into effect fairly soon into other areas of work. You can't measure impact by profit alone so I don't find the article particularly interesting anyway.
Maybe it's also rate-limited by the fact that it's not actually a silver bullet and actually needs time to find the proper place for its application as well.
Only a few curmedgeons at this point are claiming there is nothing, nothing whatsoever, to AI.
The question is really about the fate of the current build out. For those of us who have extensive experience with AI, which is to say, those of us here, because we are probably the only people not only using it day in and day out but potentially using multiple sessions of it all the time, I think it's not so surprising that it isn't necessarily immediately radically transforming every industry everywhere. Even as it has rewritten the minute-by-minute process of doing my job entirely, it hasn't actually changed my job all that much at the higher level and it's hard to assign much more than a 2x or a 2.5x overall improvement to it once I account for all the Amdahl's Law issues that come up from the parts that have not been so amenable to acceleration.
And without unit testing, continuous integration testing, pull request review, the fact that the AI is building code which is itself a highly reviewable artifact rather than just doing whatever directly, and all the many many guard rails we've built up in our industry that happen to be extraordinarily amenable to use by automated systems, I still wouldn't trust AI all that much for a real task. They still bang into those guard rails a lot, and still make what I consider to be deeply suboptimal decisions far too often to take my hands off the wheel. With all the stuff we have in the programming space they are a game changer, but in a field that doesn't have anywhere near that much padding and safety nets I'd be way, way more hesitant.
I think there's also something to be said about how the largest private AI enterprises are themselves mostly just software companies. There is enough widespread dogfooding of the product that the company staff are not just spending a lot of time and energy in improving how their product is specifically adapted and trained for better code but they are also intimately aware of those limitations, driving how they think about and implement their own product.
I don't think that translates well for other fields at all. You can throw an engineer or an MBA-brained exec at a business problem all you want but until they start using the products, they're not going to be effective at guiding its development.
This seems like a poor comparison, for any given company, benefitting from personal computing in 1987 was a significantly harder task than using AI tools in 2026 is.
The first requires a fundamental shift in not only how you work but expensive and complex infrastructure changes from the company itself just to make it possible (the first Mac for example would cost about $7000 per person in today’s money and would be outdated in maybe a year or two), an office building of The latter requires asking a subscription service to do a task for you on basically any computer connected to the internet.
I don’t know about you, but for me materially useful AI costs around $10,000/mo in tokens. But it is very useful.
Claude subscriptions help, but only if I come in with the expertise to build the entire system around the model and I don’t mind the whole thing grinding to a halt when my session limit is saturated. Anyone selling the system off the shelf is going to be charging API rates. Computers were cheap in comparison.
I can’t speak for everyone, but I would intuitively venture to guess there would be some greater than marginal productivity benefits from say, a $200/m Claude sub for each employee in many organizations outside of tech.
Perhaps they’re all on the $20/m plan which is useless?
That would be speculative, it should only be added in the hypothetical time the productivity actually goes up. Even then, it should be some years in and with third party sources like all of Wikipedia.
And at that point, there was no much productivity gained ... and not nearly as much debt as now. The industry was also not such a huge part of stock market, GDP, etc.
One industry example.. We have been giving students computing devices for years with the cost of educating students going up year after year.. Yet scores keep dropping.. No productivity gains here.. I do not think these results are unique to the education sector.. There are probably alot of businesses where nothing much will change except slight efficiencies.. not enough to move the needle..
This is an important observation. We've had 50 years of personal computing, but is there any evidence that computers in any form ever improved the academic performance of students?
AI certainly seems to be making students much less competent.
"AI does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness, all of which shape reputational risk and ultimately drive revenue growth and profit margins."
For a sec I thought he was talking about private equity.
We can use a non-deterministic system because we have test suites that act as validation gates.
Other industries don't, so they have no way of knowing if the AI is bullshitting them outside of manual validation, which makes it more expensive and slower than doing it yourself.
For those who won't act ethically or at least reasonably, regulation is the answer. We're not too far away from flipping the congress into something far more willing to do just that, especially with anything connected to Elon Musk.
No signs of AI payoff outside the tech sector...yet. I think it's partially because our harnesses are generally much, much better. And engineers seem to be more deft at figuring out where AI provides real benefit and where it needs a watchful eye.
I'm seeing non-engineers use AI to send more emails, have more meetings, create more documents, and create more dashboards. So much of the busy work that happens in corporate environments was never that valuable in the first place. A firehose of busy work doesn't move any needles that matter.
The only issue is that the benefits would be more enmeshed with other gains in other industries.,
Heck, even in Software Development, the real benefits are not with mass labor elimination resulting in cost savings but with increasing available throughput. Technology if used to the fullest potential should be able to expand the proverbial pie., which is where the gains are!
That is not to say people won't try to save cost by eliminating labor due to AI., which is super shortsighted and misses the forest for the trees, but the overarching goal should be to increase the quality and quantity of deliverables, as demanded by the market.
Here's the data for US personal consumption expenditures which comprise 2/3rds of the GDP:
U.S. PERSONAL CONSUMPTION EXPENDITURES (Q2 2026 BEA DATA)
==============
(In USD Billions)
TOTAL PERSONAL CONSUMPTION EXPENDITURES $21,993.2 100.0%
SERVICES $15,131.8 68.8%
- Housing and Utilities $3,801.2 17.3%
- Health Care $3,551.6 16.1%
- Financial Services and Insurance $1,689.8 7.7%
- Food Services and Accommodations $1,495.2 6.8%
- Recreation Services $819.2 3.7%
GOODS $6,861.4 31.2%
- Nondurable Goods (Subtotal) $4,470.6 20.3%
* Food and Beverages (off-premise) $1,360.0 6.2%
* Clothing and Footwear $510.0 2.3%
* Gasoline and Energy Goods $445.0 2.0%
- Durable Goods (Subtotal) $2,390.8 10.9%
* Motor Vehicles and Parts $786.5 3.6%
* Furnishings & Household Equipment $529.6 2.4%
==============
There will be some sectors that will benefit from AI more than others, and perhaps AI can take 20bps-100bps of spend in each of these categories. Overall, that'd be 2-3% of the overall figure equating to $400-600bn of TAM within the US. Apply your favourite return thresholds and investment time horizons here.
As the electricity analogy is often provided, the total revenues for retail electricity sales were around $514bn (EIA) in 2025 in the US. I'd argue that electricity is far more necessary and productivity enhancing than AI as it allows vast amounts of energy to be transported very conveniently to numerous end-points - no economic activity occurs without energy.
There are bigger sectors of the economy out there.
I think you could say the same thing of tech and the economy prior to AI.
No Signs of Tech Payoff Outside the Tech Sector. Manufacturing, basic science, education, etc languished while Big Tech reaped record profits and achieved stock heights for the last 15 years.
Hell, on top of that (i.e. the confounders you mention are very very apparent everywhere,) the challenge of folks understanding how to use AI to be more productive.
Even last week I had a case where, I was reviewing code that I know the dev was vibe coding, and honestly was good up to that review.
Buuuuuut, for the stuff I reviewed that day, there was some base-class inheritance use that was so ugly that my initial teams message to the dev was along the lines of 'What crack was Copilot smoking on that one'.
Turns out, they had asked some other random-ish questions in the chat they were doing their actual work in, and those questions more or less polluted the context.
Sure, I was able to guide them for how to not do that in future (i.e. keep unrelated questions around the work in question to a separate chat, etc.) but this was also a dev that is normally both very good, but also overworked semi-related to the general 'AI will save us on dev costs thing'.
I can't help but ask whether, for the even less-technical people that are told 'yeah you lost two colleagues but here is Copilot/Claude/etc' and they don't even get enough advice to have it actually increase productivity versus chasing down gremlins in the actual work.
Just talking about salaries, revenue, and stock IN AMERICA.
Doubtlessly Chinese manufacturing benefited tremendously from those advancements. But my point is the headline whinging about AI in America would also apply to Tech in America.
I'm a dev not a video or 3D guy, but I'm looking at people I used to know and also some youtube doom-scrolling, the blender MCP, cinema 4D MCP, and seeing that it's a big impact
We're not seeing full automation for things that ought to be automatable now.
Customer service chatbots still aren't very good. Salesforce ought to be taking over entire marketing operations, not just supporting sales but replacing it. The medical office side of insurance ought to be automatic now. The pick and place side of warehouse operations should be robotic. Shelf restocking in stores ought to be robotic. All of those things have been demoed, but didn't go to production.
'a.i' is unlike software before - where it was build then deploy with lack of customization except of a few high ticket items such as SAP, Workday etc.
for 'A.I' to work for the rest of industry outside of software - you need high touch customization.
We’re heading into corporate budget season. Company after company is slashing their AI budget and the head cut wagons are circling around all the various “AI lab” teams that every company invested in two years ago with zero ROI to show for it. All while massive debt piles up to build data centers despite little evidence any of this is coming close to returning the ROI needed to justify any of it.
I would say the driving industry will get up-ended, but we'll see. It may just be workers get paid less because they only babysit the truck during very complicated times.
Robots still seem far off. Its just like war, air superiority might make the war winnable but its the boots on the ground that are needed to actually take the territory.
You won't get magical productivity gains from a workforce heavily conditioned to following orders and a pre-packaged process (the Henry Ford model and its various bastardizations). In order to extract value from an LLM, you have to be able to think and ideate so that the inputs are of higher quality.
It should be of zero surprise that the rush into AI has yielded limited results (if we're honest, both in and outside of tech)—you won't get a bunch of artists or creative thinkers out of a pliant, ironically agentic labor force. That's why most people say "it doesn't work:" it's because they lack the skills and knowledge to use it properly (but corporations want to delude themselves into thinking they have a genie in a lamp).
Many will fall for the road to be paved. This is the most advanced discovery for humans to date. There is no doubt it will succeed and is here to stay. It is still new (gold rush) and deregulated. A mine field for govs and businesses. We may see to big to fail crashing (.com bubble style or worse) but no one will ever stop AI.
Let's not forget that the gold rush saw a very small percentage if people who got rich, to those who lost money. It was the people selling a pickaxe and a dream who did well.
Indeed well said, but gold is adopted worldwide now and many commerce it daily. I see it as a rewardive opportunity, some expect economical benefits, some others simply health and general knowledge, Me... breakthroughs :)
Also, the woman who has an AI child with Gemini said not to support artists who hate on AI, and I agree. We must crush the haters before it's too late, for they are irrational and will be left behind as we go vertical any second now.
seriously though: nothing will stop the sky from being blue, anytime soon. How I know this? Because we don't constantly talk about how the sky will always be blue. If you got the goods, you don't need the sales pitch.
Love it when some write nonsense just to be heard or had a bad day, maybe simply ignorance!
Fascism comes from the latin word Fascia, which has nothing to with this topic. It was implemented later on as political philosophy, which still has nothing to do with this topic.
The author of the article is trying to make a point which is correct in micro but incorrect in macro/long run. My claim is that AI is working and is here to stay (but always as a tool in our hands).
Do you understand now how ignorant you sound? You got this champ, accept this critic.
I think we are struggling to achieve payoff within the tech sector.
There have been wins with the customer facing side of the business, but the development team seems to be the biggest problem right now despite the direct & obvious applicability of these tools. I know they aren't a special unicorn edge case because I was very rapidly solving their issues before they stopped handing me things that revealed the lethargy.
Within the sector where these tools definitely work, the biggest barrier to effective adoption is figuring out a way to undermine the various ideological high grounds. I have found the most effective argument to untangle this knot to be something approximating:
Challenge:
"Why should I adopt or help to implement technology that might eventually replace my job?"
Response:
"Because refusing to won't protect your job. It puts everyone's job at risk."
It doesn't matter if you protect your job (not using AI) if the entire company goes to the bottom of the ocean as a consequence. The neat thing about this is that your job is part of the set known as "The Company" and if that ceases to exist, so does your job. Good luck suing a dissolved entity for unemployment or whatever. There is enough competition in most markets to punish anyone who thinks they can win on virtue alone.
I personally have no problem per se with AI rendering certain tasks or even jobs economically redundant (although I am sceptical about many of the grander claims). What I do find tough to take is that the benefits will accrue to companies who have effectively stolen (and continue to steal) the intellectual output of everyone based on some bullshit fair use ruling for AI training. Hopefully the latter will get walked back and mechanisms and systems will be out in place to more fairly reward human intelligence.
> It doesn't matter if you protect your job (not using AI) if the entire company goes to the bottom of the ocean as a consequence.
The vast majority of entire companies are going to the bottom of the ocean regardless of what individual contributors do.
Once you can replace nearly every job at a company, the company doesn't need to exist. Companies are just collections of experts, if AI replaces all the experts then it replaces the whole company.
So I've adopted the accelerationist mindset of: yes, this is absolutely going to replace my job, but it is going to replace everyone else's job too so fuck it I'll use it and we'll all go down to the bottom of the ocean together.
It's too early. There was a time where economist could not find productivity gains from PC products. I think I remember reading about it in the early 2000s over a decade after computers were common in business. Part of it is that companies aren't quick to throw what's work for tech that hasn't fully proven itself.
AI productivity gains are coming make no mistake about it. My guess is that we'll see them within the next 5 years.
> AI does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness, all of which shape reputational risk and ultimately drive revenue growth and profit margins.
This reminds me of the book The Redemption of Time. An alien society incapable of lying has no other way to understand deception than to study an actual human mind. The different ways that we are capable of thinking can be so hard to capture in an artificial system.
I've been doing workflow automations for many organizations for the last two years, and here is what I see:
1. every business unit and department was neglecting a best practice due to lack of investment in their organization. it has always been that way for decades and always will be that way because it isn't economically viable for it to be any other way. the workflow automations and agents we create act as digital workers that either do that neglected part, or let the rest of the organization do that neglected part.
2. the whole company always starts the year off with a bunch of goals, and by Q3 most of them are cancelled. now less of them are cancelled.
both factors mean that anyone just looking at "worker productivity" or "did we save money" is going to get the answer of no. everything is costing you the same, or the AI itself is costing you more. whatever contract I'm working under is definitely going to cost you more, if you don't cancel all those enterprise SaaS seats we told you to cancel.
I see a path to savings, but I don't think that's necessarily the metric to focus on. you need to look at the whole corporate structure, why are you still doing certain things, because those pressures will just make the org do more things because nobody is empowered to say "wait, maybe this is dumb"
great. let's pop this bubble so everyone panicking can leave us devs alone with our tools. GLM 5.3 is good enough if anthropic and openai go under (which they wont).
Vibe coding as in "Make me a postgres" is poor but if you think it hasn't affected CRUD feature work you would be wrong.
Most developers don't work building low level high performance scalable systems for a large tech firms, they are making CRUD back-ends for phone apps.
This section of the industry is definitely contracting. Even more obvious is the dev tooling part of it. Dev tooling companies are disappearing or being "bought" all over the place because what might have been a product before is now an hour or two with someone with CC.
As someone in what is relatively the equivalent of Raven's Call of Duty mines compared to past work...
Yeah, 100%. Even for relatively complex scenarios (i.e. over-engineered for purpose) as long as the engineer can prompt coherently and tweak it's a big increase in productivity, at least for those people.
The gap is in things like contract houses where you might see better code on code review from the people submitting but if you actually use the tools, you know that the hours billed are even more inflated than they might have been before.
I know of a brake parts factory outside dayton ohio who had the Microsoft folks come through with promises well over a year ago now. They've already overspent and had to pull back their AI budgets when it made more work than it solved.
> before most companies even realize it's an option
Couldn't be further from the truth
Seven years ago (*!!SEVEN!!*) GPT was too dangerous to release and yet everything that AI has touched since then is WORSE now than it was then. "Three months" is entirely nonsense. Seven years. Seven. S. E. V. E. N. Y. E. A. R. S.
If you moved the goal posts any farther, they'd have traveled around the world and would be sitting at your back right now.
The incredible improvements have only been over the last 9 months or so. Business, especially disrupting entire job functions, takes much longer than that. As is typical with this type of change, it'll be slow until a tipping point, then FOMO will hit.
If I were to count the number of times you all have said "incredible improvements" here on this site over the last 7 years, what number do you think I'd come up with, and would that change your tune? I'll do it, if I must, to show you all how un-serious is this claim.
If anything the expectation of incredible improvements just over the horizon makes an argument for not adopting AI or at least doing so very judiciously.
Why would you spend dollars on tokens today if you expect the same tokens will get you something vastly superior in a few months, or that they'll do the same job for nickels on the dollar? Do you just hate having money or what is the problem?
The tech sector is the most software-defined and virtual of all sectors, it's natural that they are first to change. Other industries need to make capital investments to adapt to change.
You can see a similar history with electrification. A few industries went quickly, but then it took like a decade of little growth before the true benefits showed up. As an example, take a factory that was built around hyrdo-power like a grain mill. You can't just take an electric motor and drop it in in place of the water wheel. It doesn't fit, the wheel is outside, there isn't room between the mill and the wall, etc. Only when you completely rebuild a new mill building designed around the features and limitations of the electric motor do you see the major benefits.
Same thing here; it will be new factories and new companies that are designed from the ground up around AI and its benefits and limitations that really move the needle.
Cool analogy, but there's another possibility: software is uniquely suited for being boosted by LLM-based coding agents. That's not to say LLMs are useless in other domains, but they may not yield as much benefit.
Software creation tools have been worked on for decades to keep fallible human coders out of trouble as much as possible. Starting with compilers that enforce or at least encourage coding style humans can understand. Compilers embody formal definitions of language syntax. Lint and other static analysis technologies contribute by identifying problematic patterns in code. Leak detectors identify memory management problems. Profilers identify performance problems.
Few other endeavors have this much tooling that serendipitously happens to help help LLM-based agents be more useful and less error prone.
A lot of discussion around AI is very high-level and abstract. It helps to dig in and identify the characteristics that actually make a domain more or less amenable to being boosted by AI.
> but they may not yield as much benefit.
This morning, I was talking to a friend of mine, who is a marketing (not tech) guy, and he was gushing about LLMs.
He loves being able to just dump in a bucket of seemingly unrelated data, and getting actionable BI from it.
I think we're in for a wild ride. Fasten your seat belt, and keep your hands inside the vehicle...
Yeah, I think it's at the level of self-driving Excel.
Now, Excel also makes mistakes while trying to get clever - remember the biologists who had to rename several genes to prevent Excel treating them as dates - but it's also a ubiquitous business tool. And you can combine them. But equally Excel hasn't replaced every white collar job.
I don't doubt it. Earlier today I got an estimate for the revenue, profit margin, and headcount for a potential supplier of a hardware component to a product I'm working on. It would've taken a couple hours of really tedious work for me to do that with Google translate, a search engine, and various market research sites. The result is I know more what I can expect from that vendor in terms of customization and support.
But those are the kinds of estimates that I can eyeball and be more than 80% certain they are plausible. That won't cut it in a lot of domains.
marketing is one of those domains, if everyone gets better, no gets better in the end
> He loves being able to just dump in a bucket of seemingly unrelated data, and getting actionable BI from it.
* what he perceives to be actionable BI.
Well, he's had a lot of real-world success, so it's likely to be something he continues to do.
Yeah but theres also no null test here. He might have done anything and still been successful. He might have done something else more optimal and done even better. We can't really tell from a single anecdote.
So…you’re point is that only his judgment says it’s useful, and maybe he shouldn’t be using it and instead rely on…his judgement.
Too true. Even on the software side i see people blindly “claude make it faster with no mistakes” and gushing about it when the output is genuinely not good not because claude isnt good but because the user of it isnt.
We got a new a CNC machine and everyone is carving away marveling at how easily it chips away at metal when the result is a jagged mess.
most of white collar work is building mid ppt and sending emails. it was automated with GPT-3.
> Cool analogy, but there's another possibility: software is uniquely suited for being boosted by LLM-based coding agents.
Or the pretty obvious third possibility: The tech sector is the part selling the LLMs. They're not digging up more gold with their cool shovel technology, they are selling the shovels.
And whole repositories with commit history that provides step by step training on how to code.
AI could probably write decently if every scraped book had all the revisions to create it embedded within or art had every brush stroke documented. As it stands programming is the first to be trained. The rest will take some more time.
If only there were a law compiler that could tell you if your legal argument was valid.
I can't speak for everywhere else, but here in the Bay Area (California), pretty much all hospitals use AI for transcription, summarizing, and some other stuff that I'm not fully aware of (I should really listen to what my Nurses say before agreeing to being recorded).
My doctors office has started doing that and it throws the wildest crap into my chart. Diagnosis the doctor didn't mention, personal anecdotes that doc told me about their family become my family medical history. Test supposedly done that weren't.
Ive been trying to figure out some persistent si joint pain and my chart is full of me passing or failed tests nobody performed, observations nobody made and diagnosis that no doctor confirmed.
Its just so much wrongness, ive tried getting it corrected a few times when it really matters but its just so much wrong crap
Speech to text is not exactly new technology.
no, but the quality jump in the last ~2 years is immense
as is translating text, deepl/google where okaisch, llm's are on another level
most lawyers haven't tried fable and astra ultrafast yet.
those who did are rather convinced it's over, for juniors at least.
It is most impactful in tech because tech is almost nothing but languages and more so super expensive languages where one hour of an “authors” time can cost thousands with a supply demand issue so bad workers are imported from across the world.
Artists were already starving. The more physical trades are…physical not language so a plumber still plums. Maybe now they ask gemini instead of google tho.
Areas related most to…language are the most impact by large language models? The one that i think is just as impacted is education because it is also almost all language driven.
I see zero sign of any of the new mills coming around. Steel, crops, farming, none of it has changed and what little has is from…the software side or…language.
The SV world really over indexes on how much software can do.
"the tech sector is the most software defined"
AIs don't just make code they make words, images, and sounds. Hasn't worked in those markets either.
There are some things that current generation LLMs are extremely good at. They also suck for a lot of reasons and appear to be prohibitively expensive to train as "general" models.
hey are amazing tools, but every time I see a task that LLMs helped me complete in 30% less time, I see 5 more tasks that took me 30-50% longer to complete because of errors, lies, mistakes, hallucinations, context limits.
Do not tell me I'm holding it wrong.
"You can see the computer age everywhere but in the productivity statistics." -- 1987
Good reference. It is going to slowly change things in 20 years we'll look back and a lot of things will be wildly different because of AI systems. But its rate limited by institutions and momentum and human psychology and the way things change will be somewhat unpredictable with unintended negative and positive consequences.
"slowly change things in 20 years"
Yeah, it is not going to be slow. Progress compounds, everything is being built on top of fast computers and fast widespread internet connection. You are not at the beginning of the curve like you were in the 20th century where people didn't even know how to use a keyboard and we are already seeing widespread disruption in security.
Literally the entire premise of the article is that it's going slow everywhere but in software. To think otherwise is to be really myopic about how the rest of the world outside of tech operates. And in any case 20 years isn't that slow. The only things that happen faster that that will be things like adobe goes out of business. And let's say AI helps us develop medicine and robotics and better factories and all the other things that would have a big impact a lot faster.. just human trials could take a decade before you really see that much change. Will it have a wider impact faster than computing its self, or the internet? Maybe? Maybe not.
If a lot of jobs were still manufacturing jobs I would agree, but we are in a service economy that uses computers as the main tool. So I don't see how it won't come into effect fairly soon into other areas of work. You can't measure impact by profit alone so I don't find the article particularly interesting anyway.
Do you think those offering professional translation service were obsessed with AI? Or chess players? The rest of the world doesn't get to decide...
"You are not at the beginning of the curve like you were in the"
Note: on an exponential, you can always say this. People will look back on today with the same "aww cute" eyes.
Maybe it's also rate-limited by the fact that it's not actually a silver bullet and actually needs time to find the proper place for its application as well.
Computers, of course, can be criticized in the same way.
Only a few curmedgeons at this point are claiming there is nothing, nothing whatsoever, to AI.
The question is really about the fate of the current build out. For those of us who have extensive experience with AI, which is to say, those of us here, because we are probably the only people not only using it day in and day out but potentially using multiple sessions of it all the time, I think it's not so surprising that it isn't necessarily immediately radically transforming every industry everywhere. Even as it has rewritten the minute-by-minute process of doing my job entirely, it hasn't actually changed my job all that much at the higher level and it's hard to assign much more than a 2x or a 2.5x overall improvement to it once I account for all the Amdahl's Law issues that come up from the parts that have not been so amenable to acceleration.
And without unit testing, continuous integration testing, pull request review, the fact that the AI is building code which is itself a highly reviewable artifact rather than just doing whatever directly, and all the many many guard rails we've built up in our industry that happen to be extraordinarily amenable to use by automated systems, I still wouldn't trust AI all that much for a real task. They still bang into those guard rails a lot, and still make what I consider to be deeply suboptimal decisions far too often to take my hands off the wheel. With all the stuff we have in the programming space they are a game changer, but in a field that doesn't have anywhere near that much padding and safety nets I'd be way, way more hesitant.
I think there's also something to be said about how the largest private AI enterprises are themselves mostly just software companies. There is enough widespread dogfooding of the product that the company staff are not just spending a lot of time and energy in improving how their product is specifically adapted and trained for better code but they are also intimately aware of those limitations, driving how they think about and implement their own product.
I don't think that translates well for other fields at all. You can throw an engineer or an MBA-brained exec at a business problem all you want but until they start using the products, they're not going to be effective at guiding its development.
To be fair, in 1987 the unit economics weren't being subsidized by nation-state levels of debt and mandatory usage contracts.
The unit economics aren't what's subsidized. Inference is profitable.
Where do you think the internet came from?
Neither of those things. Arpanet was military and the web came out of CERN, but they were both low budget side projects.
in some kind of boring sense, sure.
but have you looked up who invented the computer? and who was funding all the first versions of the software we use? it was the US military.
...and global usage/reach was practically zero...
This seems like a poor comparison, for any given company, benefitting from personal computing in 1987 was a significantly harder task than using AI tools in 2026 is.
The first requires a fundamental shift in not only how you work but expensive and complex infrastructure changes from the company itself just to make it possible (the first Mac for example would cost about $7000 per person in today’s money and would be outdated in maybe a year or two), an office building of The latter requires asking a subscription service to do a task for you on basically any computer connected to the internet.
I don’t know about you, but for me materially useful AI costs around $10,000/mo in tokens. But it is very useful.
Claude subscriptions help, but only if I come in with the expertise to build the entire system around the model and I don’t mind the whole thing grinding to a halt when my session limit is saturated. Anyone selling the system off the shelf is going to be charging API rates. Computers were cheap in comparison.
I can’t speak for everyone, but I would intuitively venture to guess there would be some greater than marginal productivity benefits from say, a $200/m Claude sub for each employee in many organizations outside of tech.
Perhaps they’re all on the $20/m plan which is useless?
The difference, of course, is that in this analogy Commodore has to become a trillion dollar business by 1989 or the entire stock market implodes.
Came here to say this as well.
It took decades for effects of personal computing to show up in aggregate statistics.
See:
https://en.wikipedia.org/wiki/Productivity_paradox and https://dl.acm.org/doi/10.1145/163298.163309
Someone should add a 2025–2045 section in the Wikipedia page of Productivity Paradox
The decade ahead is going to be a very interesting one
That would be speculative, it should only be added in the hypothetical time the productivity actually goes up. Even then, it should be some years in and with third party sources like all of Wikipedia.
And at that point, there was no much productivity gained ... and not nearly as much debt as now. The industry was also not such a huge part of stock market, GDP, etc.
I came here to say same thing :-)
5 years ago it was the crypto bros quoting this
One industry example.. We have been giving students computing devices for years with the cost of educating students going up year after year.. Yet scores keep dropping.. No productivity gains here.. I do not think these results are unique to the education sector.. There are probably alot of businesses where nothing much will change except slight efficiencies.. not enough to move the needle..
This is an important observation. We've had 50 years of personal computing, but is there any evidence that computers in any form ever improved the academic performance of students?
AI certainly seems to be making students much less competent.
"AI does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness, all of which shape reputational risk and ultimately drive revenue growth and profit margins."
For a sec I thought he was talking about private equity.
I'm sure private equity understands it very well... That's why they profit so much, they understand what to destroy.
We can use a non-deterministic system because we have test suites that act as validation gates.
Other industries don't, so they have no way of knowing if the AI is bullshitting them outside of manual validation, which makes it more expensive and slower than doing it yourself.
> they have no way of knowing if the AI is bullshitting them
That hasn’t stopped Tesla. Their self-driving cars work by connecting the car’s controls directly to a neural network.
For those who won't act ethically or at least reasonably, regulation is the answer. We're not too far away from flipping the congress into something far more willing to do just that, especially with anything connected to Elon Musk.
right cause test suites never have issues
No signs of AI payoff outside the tech sector...yet. I think it's partially because our harnesses are generally much, much better. And engineers seem to be more deft at figuring out where AI provides real benefit and where it needs a watchful eye.
I'm seeing non-engineers use AI to send more emails, have more meetings, create more documents, and create more dashboards. So much of the busy work that happens in corporate environments was never that valuable in the first place. A firehose of busy work doesn't move any needles that matter.
Article is spot on!
The only issue is that the benefits would be more enmeshed with other gains in other industries.,
Heck, even in Software Development, the real benefits are not with mass labor elimination resulting in cost savings but with increasing available throughput. Technology if used to the fullest potential should be able to expand the proverbial pie., which is where the gains are!
That is not to say people won't try to save cost by eliminating labor due to AI., which is super shortsighted and misses the forest for the trees, but the overarching goal should be to increase the quality and quantity of deliverables, as demanded by the market.
Here's the data for US personal consumption expenditures which comprise 2/3rds of the GDP:
U.S. PERSONAL CONSUMPTION EXPENDITURES (Q2 2026 BEA DATA)
==============
(In USD Billions)
TOTAL PERSONAL CONSUMPTION EXPENDITURES $21,993.2 100.0%
==============There will be some sectors that will benefit from AI more than others, and perhaps AI can take 20bps-100bps of spend in each of these categories. Overall, that'd be 2-3% of the overall figure equating to $400-600bn of TAM within the US. Apply your favourite return thresholds and investment time horizons here.
As the electricity analogy is often provided, the total revenues for retail electricity sales were around $514bn (EIA) in 2025 in the US. I'd argue that electricity is far more necessary and productivity enhancing than AI as it allows vast amounts of energy to be transported very conveniently to numerous end-points - no economic activity occurs without energy.
There are bigger sectors of the economy out there.
I think you could say the same thing of tech and the economy prior to AI.
No Signs of Tech Payoff Outside the Tech Sector. Manufacturing, basic science, education, etc languished while Big Tech reaped record profits and achieved stock heights for the last 15 years.
Let's not forget tech was in a major slowdown announcing layoffs left and right before AI picked up
End of ZIRP, tax law changes, post-COVID inflation, and general economic malaise from federal mismanagement.
There are too many confounders.
Hell, on top of that (i.e. the confounders you mention are very very apparent everywhere,) the challenge of folks understanding how to use AI to be more productive.
Even last week I had a case where, I was reviewing code that I know the dev was vibe coding, and honestly was good up to that review.
Buuuuuut, for the stuff I reviewed that day, there was some base-class inheritance use that was so ugly that my initial teams message to the dev was along the lines of 'What crack was Copilot smoking on that one'.
Turns out, they had asked some other random-ish questions in the chat they were doing their actual work in, and those questions more or less polluted the context.
Sure, I was able to guide them for how to not do that in future (i.e. keep unrelated questions around the work in question to a separate chat, etc.) but this was also a dev that is normally both very good, but also overworked semi-related to the general 'AI will save us on dev costs thing'.
I can't help but ask whether, for the even less-technical people that are told 'yeah you lost two colleagues but here is Copilot/Claude/etc' and they don't even get enough advice to have it actually increase productivity versus chasing down gremlins in the actual work.
What? Data please.
Manufacturing does not benefit from robots and modern machines? Sciene does not benefit from computers?
I don't follow.
Just talking about salaries, revenue, and stock IN AMERICA.
Doubtlessly Chinese manufacturing benefited tremendously from those advancements. But my point is the headline whinging about AI in America would also apply to Tech in America.
It's almost like Big Tech is a net drain on everyone.
In video production, 3D modelling, editing etc, I think a lot of jobs are disappearing, with cost dropping. Not data backed, just from what I can see.
Are the results in video generation precise enough to remove the need for 3D modeling?
I'm a dev not a video or 3D guy, but I'm looking at people I used to know and also some youtube doom-scrolling, the blender MCP, cinema 4D MCP, and seeing that it's a big impact
We're not seeing full automation for things that ought to be automatable now. Customer service chatbots still aren't very good. Salesforce ought to be taking over entire marketing operations, not just supporting sales but replacing it. The medical office side of insurance ought to be automatic now. The pick and place side of warehouse operations should be robotic. Shelf restocking in stores ought to be robotic. All of those things have been demoed, but didn't go to production.
'a.i' is unlike software before - where it was build then deploy with lack of customization except of a few high ticket items such as SAP, Workday etc.
for 'A.I' to work for the rest of industry outside of software - you need high touch customization.
We’re heading into corporate budget season. Company after company is slashing their AI budget and the head cut wagons are circling around all the various “AI lab” teams that every company invested in two years ago with zero ROI to show for it. All while massive debt piles up to build data centers despite little evidence any of this is coming close to returning the ROI needed to justify any of it.
2027 is going to be very interesting. Buckle up.
My sector used it to make all the tools we wanted, then put it back in its box awhile ago. Just enjoying the fruits rent free.
Tell that to legal firms and patent guys
I doubt Excel'ification happens this fast. To make it "payoff" it needs to become the next Excel-levels of usage.
I would say the driving industry will get up-ended, but we'll see. It may just be workers get paid less because they only babysit the truck during very complicated times.
Robots still seem far off. Its just like war, air superiority might make the war winnable but its the boots on the ground that are needed to actually take the territory.
You won't get magical productivity gains from a workforce heavily conditioned to following orders and a pre-packaged process (the Henry Ford model and its various bastardizations). In order to extract value from an LLM, you have to be able to think and ideate so that the inputs are of higher quality.
It should be of zero surprise that the rush into AI has yielded limited results (if we're honest, both in and outside of tech)—you won't get a bunch of artists or creative thinkers out of a pliant, ironically agentic labor force. That's why most people say "it doesn't work:" it's because they lack the skills and knowledge to use it properly (but corporations want to delude themselves into thinking they have a genie in a lamp).
Many will fall for the road to be paved. This is the most advanced discovery for humans to date. There is no doubt it will succeed and is here to stay. It is still new (gold rush) and deregulated. A mine field for govs and businesses. We may see to big to fail crashing (.com bubble style or worse) but no one will ever stop AI.
Let's not forget that the gold rush saw a very small percentage if people who got rich, to those who lost money. It was the people selling a pickaxe and a dream who did well.
Indeed well said, but gold is adopted worldwide now and many commerce it daily. I see it as a rewardive opportunity, some expect economical benefits, some others simply health and general knowledge, Me... breakthroughs :)
Also, the woman who has an AI child with Gemini said not to support artists who hate on AI, and I agree. We must crush the haters before it's too late, for they are irrational and will be left behind as we go vertical any second now.
seriously though: nothing will stop the sky from being blue, anytime soon. How I know this? Because we don't constantly talk about how the sky will always be blue. If you got the goods, you don't need the sales pitch.
fascist thinking
Love it when some write nonsense just to be heard or had a bad day, maybe simply ignorance! Fascism comes from the latin word Fascia, which has nothing to with this topic. It was implemented later on as political philosophy, which still has nothing to do with this topic. The author of the article is trying to make a point which is correct in micro but incorrect in macro/long run. My claim is that AI is working and is here to stay (but always as a tool in our hands). Do you understand now how ignorant you sound? You got this champ, accept this critic.
I think we are struggling to achieve payoff within the tech sector.
There have been wins with the customer facing side of the business, but the development team seems to be the biggest problem right now despite the direct & obvious applicability of these tools. I know they aren't a special unicorn edge case because I was very rapidly solving their issues before they stopped handing me things that revealed the lethargy.
Within the sector where these tools definitely work, the biggest barrier to effective adoption is figuring out a way to undermine the various ideological high grounds. I have found the most effective argument to untangle this knot to be something approximating:
Challenge:
"Why should I adopt or help to implement technology that might eventually replace my job?"
Response:
"Because refusing to won't protect your job. It puts everyone's job at risk."
It doesn't matter if you protect your job (not using AI) if the entire company goes to the bottom of the ocean as a consequence. The neat thing about this is that your job is part of the set known as "The Company" and if that ceases to exist, so does your job. Good luck suing a dissolved entity for unemployment or whatever. There is enough competition in most markets to punish anyone who thinks they can win on virtue alone.
I personally have no problem per se with AI rendering certain tasks or even jobs economically redundant (although I am sceptical about many of the grander claims). What I do find tough to take is that the benefits will accrue to companies who have effectively stolen (and continue to steal) the intellectual output of everyone based on some bullshit fair use ruling for AI training. Hopefully the latter will get walked back and mechanisms and systems will be out in place to more fairly reward human intelligence.
> Hopefully the latter will get walked back and mechanisms and systems will be out in place to more fairly reward human intelligence.
Is there even any actual basis to walk this back at all anymore? It kinda seems like its a done deal and this is the outcome.
> It doesn't matter if you protect your job (not using AI) if the entire company goes to the bottom of the ocean as a consequence.
The vast majority of entire companies are going to the bottom of the ocean regardless of what individual contributors do.
Once you can replace nearly every job at a company, the company doesn't need to exist. Companies are just collections of experts, if AI replaces all the experts then it replaces the whole company.
So I've adopted the accelerationist mindset of: yes, this is absolutely going to replace my job, but it is going to replace everyone else's job too so fuck it I'll use it and we'll all go down to the bottom of the ocean together.
Payoff doesn’t matter. The future potential (payoff) is what matters.
And we’ve seen tech sector payoff?
I doubt it.
People are, on average, too stupid to make good use of AI to the point where you can actually gain massive productivity.
The pointy haired bosses are going to do their usual approach: force people to work more, claim success, move on.
all I've seen are large tech sector bills for token usage and investor capital trying to make ai happen
Well, writers can use it to speed up the draft review process. Not a big financial deal, but it's a 'payoff' for the craft.
Those with the payoff might not be sharing.
It's too early. There was a time where economist could not find productivity gains from PC products. I think I remember reading about it in the early 2000s over a decade after computers were common in business. Part of it is that companies aren't quick to throw what's work for tech that hasn't fully proven itself.
AI productivity gains are coming make no mistake about it. My guess is that we'll see them within the next 5 years.
> AI does not understand accountability, trust, culture, unwritten rules, ethical trade-offs or fairness, all of which shape reputational risk and ultimately drive revenue growth and profit margins.
This reminds me of the book The Redemption of Time. An alien society incapable of lying has no other way to understand deception than to study an actual human mind. The different ways that we are capable of thinking can be so hard to capture in an artificial system.
Trickle-down AI economics here we go! Suuurely it will work this time, any minute now don't worry about it.
Pretty much the same thing inside tech, I love AI for my personal projects but the productivity return in my dayjob has been slightly negative so far.
There is two ways of looking at this - one is that things are overhyped.
The other is that there is so much more value to be extracted before this bubble burst.
Not sure which I believe, objectively. Subjectively, inside of tech, it certainly feels like the former.
For AI to be properly integrated into businesses it would need to be enabled by people who would shortly after become irrelevant.
This may well remain a "true communism/socialism has never been tried" situation.
Related:
Only 13% of companies see real returns from AI, survey finds - https://news.ycombinator.com/item?id=50207881 - October 2026
13% of all companies, or only those that invested in AI. And if invested, by how much?
is this 13% of leprechauns become elves?
I've been doing workflow automations for many organizations for the last two years, and here is what I see:
1. every business unit and department was neglecting a best practice due to lack of investment in their organization. it has always been that way for decades and always will be that way because it isn't economically viable for it to be any other way. the workflow automations and agents we create act as digital workers that either do that neglected part, or let the rest of the organization do that neglected part.
2. the whole company always starts the year off with a bunch of goals, and by Q3 most of them are cancelled. now less of them are cancelled.
both factors mean that anyone just looking at "worker productivity" or "did we save money" is going to get the answer of no. everything is costing you the same, or the AI itself is costing you more. whatever contract I'm working under is definitely going to cost you more, if you don't cancel all those enterprise SaaS seats we told you to cancel.
I see a path to savings, but I don't think that's necessarily the metric to focus on. you need to look at the whole corporate structure, why are you still doing certain things, because those pressures will just make the org do more things because nobody is empowered to say "wait, maybe this is dumb"
great. let's pop this bubble so everyone panicking can leave us devs alone with our tools. GLM 5.3 is good enough if anthropic and openai go under (which they wont).
Yeah because vibe coding has only really worked for 3 months. Give it a chance!
It'll probably be a couple of years before most companies even realise it's an option.
Vibe coding as in "Make me a postgres" is poor but if you think it hasn't affected CRUD feature work you would be wrong.
Most developers don't work building low level high performance scalable systems for a large tech firms, they are making CRUD back-ends for phone apps.
This section of the industry is definitely contracting. Even more obvious is the dev tooling part of it. Dev tooling companies are disappearing or being "bought" all over the place because what might have been a product before is now an hour or two with someone with CC.
As someone in what is relatively the equivalent of Raven's Call of Duty mines compared to past work...
Yeah, 100%. Even for relatively complex scenarios (i.e. over-engineered for purpose) as long as the engineer can prompt coherently and tweak it's a big increase in productivity, at least for those people.
The gap is in things like contract houses where you might see better code on code review from the people submitting but if you actually use the tools, you know that the hours billed are even more inflated than they might have been before.
I know of a brake parts factory outside dayton ohio who had the Microsoft folks come through with promises well over a year ago now. They've already overspent and had to pull back their AI budgets when it made more work than it solved.
> before most companies even realize it's an option Couldn't be further from the truth
The problem with this argument is that people have been saying "it just recently got good" for about two years now.
Seven years ago (*!!SEVEN!!*) GPT was too dangerous to release and yet everything that AI has touched since then is WORSE now than it was then. "Three months" is entirely nonsense. Seven years. Seven. S. E. V. E. N. Y. E. A. R. S.
If you moved the goal posts any farther, they'd have traveled around the world and would be sitting at your back right now.
We're a little early to be calling the game yet.
The incredible improvements have only been over the last 9 months or so. Business, especially disrupting entire job functions, takes much longer than that. As is typical with this type of change, it'll be slow until a tipping point, then FOMO will hit.
If I were to count the number of times you all have said "incredible improvements" here on this site over the last 7 years, what number do you think I'd come up with, and would that change your tune? I'll do it, if I must, to show you all how un-serious is this claim.
If anything the expectation of incredible improvements just over the horizon makes an argument for not adopting AI or at least doing so very judiciously.
Why would you spend dollars on tokens today if you expect the same tokens will get you something vastly superior in a few months, or that they'll do the same job for nickels on the dollar? Do you just hate having money or what is the problem?