Not anymore! Well, if you're like Elon and already taking down the bottle of Cuatro Comas from the high shelf, the economies of scale will continue to work in your favor.
But one of the really neat things about AI is that there is no limit in sight to the scaling incentive. More compute will always get you more: more training, more inference, more parameters, more capacity to build more and better models, more spare capacity to run the slop your models have already built to generate the slop that will succeed it. Back in the dot-com days, or even the "big data" days, you wanted to scale up rapidly but there was a limit: there were only so many customers and they could only produce so much data you could only ingest so fast. In the late 90s, one of the world's most trafficked sites, ftp.cdrom.com, ran on a (single!) dual-processor Pentium Pro system. That was just serving files, and there was certainly room for more CPU oomph to provide more sophisticated services to a huge customer base. But once those customers were served, more compute, storage, and network capacity didn't buy you enough to justify the capex. That is emphatically not the case with AI, and so the incentives for the AI companies are to buy as much compute as they possibly can. What this means in practicing is pre-purchasing capacity at the semiconductor fabs to manufacture chips exclusively for you, and there's only so much of that capacity in the world. Trillion-dollar companies can easily outbid the entire consumer market, and so the incentives for the fabs are now to sell to AI companies at the expense of the consumer market. That's why you're seeing memory prices go through the roof. Modularized RAM for end-user PC builds will soon go the way of the CRT: it will cease to exist as a market product, it won't be manufactured anywhere by anyone. GPUs, CPUs, and storage will soon follow. The only devices end users will be permitted to purchase are all-in-one integrated devices, with CPU, RAM, GPU, storage, and networking either integrated in-chip or soldered on, and they will have just enough capacity to connect to the cloud services the user wants most to use. Most likely, you will be permitted a subscription to such a device, with automatic hardware upgrades at periodic intervals supplied by the manufacturer. If your subscription lapses the device bricks itself. Almost certainly, the OS will be locked down, with no end-user option to install a different one or even run unapproved software.
If reasonably powerful computer hardware for end users exists in this future, it will be available from a single company: Apple. Only they have the leverage to prevent ~100% of manufacturing capacity from going to high-roller, big-tech firms.
> Trillion-dollar companies can easily outbid the entire consumer market
I don't think this is true. I think prices are rising at the consumer and prosumer level because that's what's required for the mass market to collectively outbid the handful of trillion-dollar companies, at least for the limited share of production they can sustainably demand. This process can continue pretty much indefinitely.
> But one of the really neat things about AI is that there is no limit in sight to the scaling incentive.
How you can be so confident? I can imagine there is some limit and with each scaling iteration gain you achieved will decrease so that further iterations would be more and more look pointless
I'm sure a limit will come around eventually. But plans are afoot to build city-sized data centers, and even then that's not enough to sate the AI superscalers' ambitions, hence Elon's talk about putting data centers in space. This is a level of compute scaling unheard of in our lifetime, and we're still a long, long way off from AGI. So while the juice may theoretically not be worth the squeeze at some point, with the current capacity we have there is no end within sight to the incentive to build more. It will take a number of years at least, and who knows how much environmental/economic destruction, before the dropoff in return on capex begins in earnest.
Recently I fired up Gemma4-26B-A4B on my 8-year-old PC... and it ran surprisingly well!
But I am going to need a much beefier machine to get it to the point where it can do any but very trivial dev tasks acceptably fast, and I'm going to need a much beefier model, perhaps one not so aggressively quantized, to keep it on task without the wheels completely falling off. Already we're talking serious money outlay, perhaps still within my programmer salary to accommodate, but just barely. And we're not even where near the performance characteristics a frontier model can support.
DGX Spark runs this sized model (I personally like qwen36moe better than gemma4moe) at speeds fast enough for interactive coding sessions. Algorithmic advances like DiffusionGemma ~4x token gen speeds (https://deepmind.google/models/gemma/diffusiongemma/)
Their apparent inability to get the basics right makes me severely doubt their claims of self-improving AI. The humans at Anthropic wouldn't know improvement if it landed on their lap and started twerking, and AI cannot do a job without strong human intervention into what the goals and guardrails actually are.
I'm kind of reminded of when Microsoft claimed it took a team of Ph.D.s to write a terminal application that updated at 60fps, and then Casey Muratori did it over a weekend. And this was before AI was writing code in earnest; when LLM-induced brainrot really sets in, civilization is in for a world of fresh hurt: lots more generated code, almost all of it garbage. And the promised AI crossover point where it becomes AGI, or indistinguishable from for software design purposes, recedes into the infinite future.
Reading this it occurs to me that this timeline may be moving toward a future where garbage software, garbage information etc. have become the norm for so long that the number of people who can distinguish trash from quality, or signal from noise, has become negligible.
A true era of ignorance, looking like an ocean of nonsense in which no one can really navigate as it is ungrounded in reality.
Idiocracy presents a naively gentle positive version of such future but there are many darker ones possible.
> I'm kind of reminded of when Microsoft claimed it took a team of Ph.D.s to write a terminal application that updated at 60fps, and then Casey Muratori did it over a weekend.
This is the same Microsoft that is now rewriting the TypeScript type checker, parser and its developer tools in Go after realizing that the bottleneck was...the performance of TypeScript itself, which is a basic compiled vs interpreted difference.
> And this was before AI was writing code in earnest; when LLM-induced brainrot really sets in, civilization is in for a world of fresh hurt: lots more generated code, almost all of it garbage.
Some folks using LLMs wouldn't realize why it makes zero sense to use TS / JS for building performant and optimal applications. This is why people were experiencing significant rendering bugs in terminal apps (they are not designed for that) and slow starts with Claude Code, which was completely vibe coded with Ink.
If you don't understand the basic fundamentals of what you are working on with LLMs and bugs are creeping up left and right, then you are just sinking in your own comprehension debt.
I watched the video and I wish I could get those 13 minutes of my life back.
He could have done it in 13 seconds instead of 13 minutes: "Anthropic is lying about the effectiveness of agentic loops because there's this one screen flicker bug in Claude Code that took a year to fix."
Yeah, like when United Airlines claims a plane can fly 300 people 6,000 miles they are lying to you.
I can prove they're lying to you because people have been complaining about uncomfortable seats and flight delays for literally decades and those issues still aren't fixed.
I write little bits of code then test them. If the screen flickers i wouldnt continue until it is solved. If ive missed it and have to hear it from (a) user(s) i would kinda like to fix it imediately. Not always possible but truly annoying things have priority over new things.
If i had unlimited developers at my command. All many times as fast as me.... how can i keep the problem? It would take some huge effort to keep.
The unlimited number of devs would talk about it an unlimited number of times. Not fixing it would be very expensive that way.
He spent that much time and you still misunderstood the direct message and missed the subtext.
The lie is coding is solved, the proof is they had an outstanding coding issue they were working on for over a year while saying coding is solved. There’s a great number of other issues with their own software that disprove their premise, but you only need one counter example to disprove something.
And because you missed it, the subtext was they want you to use loops not because they work but because they burn lots of tokens thus making them more money.
One of the things I like about it is he had to create a little front end to display the game, mirroring actual COBOL practice. Until the 80s or so, COBOL didn't support meaningful terminal I/O in its own right; it was all batch. If you were a mainframe dev and wanted to do terminal interaction, you either had to write your own routines for that in mainframe assembly or use something like CICS, IBM's application server which provided its own terminal handling and transactional database routines accessible through a language extension which got swizzled into regular COBOL by a preprocessor. Creating a layer outside of COBOL to do the things COBOL was deficient in, and using COBOL's regular I/O to communicate with it, is peak mainframe-era engineering.
Other solutions to the same problem existed; there was one called InterComm, which lacked the preprocessor and required you to reserve a shared area of memory and write messages to InterComm directly into it. These days there's KICKS, an open source library API-compatible with CICS, aimed at the sort of person who faffs about with old software on Hercules.
That just gave me flashbacks. I wrote code on some ancient stuff that operated like you describe InterComm. Haven't done that in years, and I don't think much of anyone uses those systems anymore (Harris H100).
I don't know what they're complaining about. AI has freed us from the drudgery of craftsmanship, letting us focus on the important stuff—managerial and administrative work!
(There's a reason why I call it the MBA's stone. It transmutes all knowledge work into a problem of management.)
AI is the new cloud. There's no market for people or companies who aren't committed to it. If you're a dev who refuses to use AI, no company will hire you; and should a company decide not to use AI they will have a hard time retaining devs (and they will need more devs). Their investors and big-ticket customers will also think twice before signing off on major commitments.
So yes, use AI. Don't nitpick the costs and benefits. The world is headed this way; if you want to develop software for a living and afford to eat, you need to be too.
> and should a company decide not to use AI they will have a hard time retaining devs (and they will need more devs)
Need more devs? Why? If a company was being profitable just fine prio AI era, they will still be profitable if they decide not to use AI. Shipping crap faster is not a formula for success. Shipping quality faster? I prefer shipping quality at a good pace
Will you though? I see this repeated, but I’ve almost never changed products because one has 10x more features than another.
I usually buy and use products that are simple and effective, and that get out of my way to do the thing that I want to do.
For email, I’m a happy customer of Fastmail and I’ve been paying them for years. I don’t care if they ever release a new feature and I’d never switch away from them to a competitor that’s less stable but does more. They release improvements slowly but they are very stable. But I would switch away from them if they start shoving AI into things or delivering subpar features that make my email worse.
For healthcare related websites, I can already see my test results, schedule appointments, and communicate with my doctor. What more could an AI-driven medical platform give me that makes my life better?
For maps — I unfortunately had to move away from Apple recently when they added Ads. So I’m mostly just using OpenStreetMaps. I could see AI improving the OSM functionality by updating the app (OrganicMaps) routing algorithms and such, so there is room for growth there, but it’s not that massive.
Can anyone offer features that Uber can’t now due to LLMs? There are a bunch of local Uber competitors but uber wins because it’s easy and there aren’t major features to differentiate there.
Do you have examples that prove that delivering a bunch of features really fast is going to steal customers from something?
your personal experience is clearly in the techie bubble
ai is more than delivering features fast (thats probably one of the lowest priorities for companies)
right now its a race to automate work, especially back office. companies already are seeing 10M+ in savings and revenue growth and we're barely starting. workflows in sales, outbound, gtm, marketing, eng, operations, compliance, kyc etc
consumer is a different beast, consumers want convenience which has already been hyperoptimized and the big consumer cos run on network effs instead of features
> right now its a race to automate work, especially back office. companies already are seeing 10M+ in savings and revenue growth
What? Where? Citations please. We're seeing big companies massively stall in all traditional sectors except AI which is a irrational market built off hype.
ai companies know how to use the tools internally, not surprising
there are no citations yet because this is going on behind closed doors, if you know you know. we'll start seeing it in the financials of companies soon. alternatively you can look at the revenue growth of applied ai companies
> So yes, use AI. Don't nitpick the costs and benefits. The world is headed this way; if you want to develop software for a living and afford to eat, you need to be too.
It's really saddening to see software engineers throw out all critical thinking and innovation out the window to behave like sheep and follow the trend line. The industry was trailblazed by people that refused to do just that and the same is going to be true in the future.
Yes but most people aren’t trailblazers, they’re horribly average and just want to get by. This includes me.
This industry is hostile. You need a self preservationist mindset if you want long term success. Or, at least, it feels that was for someone who isn’t absurdly talented, wealthy, or connected. So, for now, we put our head down and be good little cogs.
I think of it this way. In a company with a good culture, I will build a rapport with my management and give my honest opinions. In a company with a bad culture, I just nod along and say “uh huh yeah yes sir what a great idea!” Because I know that’s it’s gonna get pushed through no matter what I say, and my opinions will only serve to hurt me.
And, right now, tech, or even America, is like one big company with a rotten, rotten culture.
It's much more than that. There's plenty of market for people and companies who don't specialise in cloud software. The market for programmers who don't use AI is going to be more like the market for programmers who don't use compilers.
reply