I've always been interested in "building". Programming was always just the means to do so.
I got into it at a very young age, but the syntax and the languages were just the means to an end of building and creating. I'm academically interested in algorithms and data structures and distributed systems architectures, but I have no desire to actually code those things by hand.
You're probably not hand-writing your software in assembler.
Next to nobody will be writing code in 2050. It was always just a stepping stone.
And all of these things will be solved one by one.
It's astounding to me that people can see coding get solved and not think every single one of these tasks won't be solved too.
Why do you not think these things aren't going to be completely automated? What makes these tasks special?
Fable and Astra can one-shot video games with compelling novel game loops. They can do systems programming, distributed systems, robotics. I haven't found a weak point.
Seedance 2.5 can make video better than the manual labor of VFX artists, 3D artists, and animators.
Nano Banana and GPT Image can do a better job than graphics designers.
LLMs just solved a Millennium Prize Problem, and there are probably more that will fall in the coming weeks.
Just wait. All of these things will be solved.
There is no "stopping point".
Edit:
Don't anticipate that 2036 will look anything like 2026.
Will Smith spaghetti doesn't stay that way forever. Trillions of dollars will be spent on solving these problems. They will be solved.
May the iterative loop of adding new axes to evaluate on be a natural, healthy progression, instead of needing to frame it as an us-them problem?
If you value humans intrinsically, this is necessarily the loop that will converge. I don't think humans have deep intensional a priori knowledge of the structure of reality. If we did, then we wouldn't need tools like AI because we'd be a superset of that. We can only observe and judge.
If we don't value humans, then sure, I think AI is at the point where it can kill all humans (conditional on sentience and resources etc). Two ways to solve a problem - solve the problem, or eliminate the problem statement. Plenty of easier vectors to eliminate the "problem statement", than say, try to solve problems such as making human life better. If you do value the latter though, there will necessarily be human judgers. That's how it works.
But lets be fair, if an expert would use AI today to build something with this, I would feel a lot more confident than not doing this.
I would start with the base architecture and add all the guardrails for a distributed system, i might even go so far to leverage the math skills of a frontier model like fable or astra. I would for sure have the proper budget for using Fable/Astra.
Tell that to the mountain of failed AI slop games on Steam! As a game dev, building compelling, fun games is not even something humans are good at doing consistently. The AI can build the tech, but it can't make something 'fun' yet (unless your bar for fun is simply that a tool created a thing).
I think the usage of solved is silly. Things get automated, they basically never get ‘solved’.
Will all these things get automation? Yeah sure. But the idea that they will be perfect automated solutions applicable in all cases is just marketing, it’s not reality.
> Fable and Astra can one-shot video games with compelling novel game loops. They can do systems programming, distributed systems, robotics. I haven't found a weak point.
Why does it matter that homes "aren't clean" while Google is? That's silly. There's going to be the same level of demand no matter what - the same mix of coal and nuclear inputs will happen regardless of where the outflows are. It's fungible.
More demand will in time result in an increase in supply. That's a good thing. We'll be building a lot more, and it's long overdue.
Google is buying the output because they are increasing demand by half a nuclear power plant's worth of power. I actually think it's pretty reasonable to pass a law demanding they build an entire nuclear power plant here.
"The agreement with Fortum provides long-term financial certainty for the Loviisa plant, which currently generates about 10% of Finland's electricity.
Fortum said Google's commitment would support an investment programme aimed at extending the life of the nuclear station and increasing its generating capacity."
Long-term PPAs are what you usually want because they are bankable and give you solid demand on which to expand. Which is literally what they're talking about doing.
If I build a house and the demand goes up I don't have to build the solar panels (and everything else involved) to cover it, I pay the utility company to build that based on how much I'm needing to use. Well, normally I pay for whatever they think will make them the most money to deliver that much electricity - Google is specifically paying to make sure the utility company sees it as increased demand and revenue for more nuclear.
Governments are extremely bad at calculating how much something "should" cost. Profits are revenue minus expenses. "Expenses" are revenue to someone else. Whenever you try to cap "profit" of some entity that isn't actually subject to competitive pressure, the next thing that happens is that the people running it start colluding or taking kickbacks from any of the people whose revenue constitutes their expenses.
Because then the utility company's expenses go up, which they would normally not want because it reduces their profit, except that then they can say that their profit isn't over the threshold and use that to increase revenue (i.e. raise prices) and still make the same profit but now also get the kickbacks under the table or help out their cronies by overpaying them.
It also in general deprives them of the incentive to improve efficiency, because reducing waste lowers expenses, which in a normal system you have the incentive to do because it means you make more money, but not if profit is a fixed amount defined by the law regardless of what you do.
Which is to say, it's much better to find a way to narrow the utility monopoly as much as possible and subject them to competitive pressure as much as possible, than to try to pretend that you can limit the profit of an entire monopoly supply chain by declaring that a single entity in it can only have a specific amount.
Do those problems exist as much if the excess was forced into them investing into a certain type of power generation? It seems less "heavy handed", since it's a proper investment for them.
I enjoy a particularly nasty utility monopoly where I am, in a dark blue state, where there is exactly zero political interest in breaking any monopolies. Any competition would probably have to be forced at a federal level. I think less bribing of politicians would happen if you tried to make them do something they were basically already going to do.
> Do those problems exist as much if the excess was forced into them investing into a certain type of power generation?
The premise of "excess" already has the problem baked into it.
Suppose an efficient company could provide service for $100 with a $10 profit, but regulators have no practical way of establishing this. The incumbent is instead providing service for $200 with a $10 profit, but then receiving $30 in kickbacks under the table or something of equivalent value not being accounted for in the official books from whoever is now getting $190 instead of $90.
Whereas getting them to invest in a particular type of generation was never hard because you can just give them a financial incentive to do it. If they have to pay a carbon tax out of their profits, which generation types are they going to add?
> If they have to pay a carbon tax out of their profits, which generation types are they going to add?
I see. Thanks for understand! How do you motivate them to actually add capacity? With a monopoly, it seems that adding supply for something people have to use could just reduce rates/profits.
At this point with all the free-market-hand waving going around, it feels like someone telling me that it's not urin raining down on me, just sparkling trickle down economics.
King Richard II would have killed for an iPhone, PS5, air conditioning, and Uber Eats. He'd swap places with you, and he was king. You don't realize how awesome you have it.
Just a hundred years ago your ancestors were (probably) eating rotten meat and drinking unclean water.
You’re assuming a large enough supply that can absorb demand. This is true in China but not generally in the West.
If supply was just in time constructed so you maybe have only like 10-30% extra vs peak, the 50% of a nuclear power plant shifts the preexisting non-Google demand onto the non-nuclear sources as the plant can no longer meet its non-Google demand. It’s generally no different than if Google acquired that electricity from the available mix rather than the plant directly.
The thing it gives them is the ability to meet their individual net-0 targets (and thus government incentives) - a local optimization at the cost of a global net negative (at least as far as carbon electricity sources goes)
This is a case where the supply was about to be gone and Google provided funds to keep it operating (the original nuclear reactor began operation in 1977 and it now needed EUR 700M to extend the service life). And one can probably guess that this supply just wasn’t cost competitive in the first place, but Google increased the demand so much that it now becomes cost competitive.
Source? The provider had requested to extend the lifetime of that particular reactor to 2050 2 years ago (something they’d started considering 8 years ago). Now maybe they did this contingent on expecting this Google deal to materialize, but it’s not clear to me these Google funds are strictly the only reason they’d keep operating the plant.
> Currently, about 80% of the projects required to extend the service life and investments of EUR700 million still lack an investment decision. Fortum said the PPA with Google will generate a predictable revenue stream, enabling it to complete the life extension investments at Loviisa.
I think they requested the operating license extension before the actual decision to extend was made.
People downvote me when I say we're back in the "Geocities era" / "indie web" of the 90's and 00's, but LLMs have enabled us to build cool side projects to completion and give them love.
This is 100% a good thing.
I love this website. It's fantastic. I don't care if it's AI. The world is better for it existing.
I agree that this is an example of interesting and good AI usage. The person who made it obviously put thought into it. Unfortunately a lot of vibe coded sites do not have that same level of care
People were already building cool side projects to completion and giving them love before AI.
You get downvoted for being weirdly ecstatic about AI "enabling" things that were already possible, while not caring about the human element that AI lacks.
People that say "things [...] were already possible" don't dream enough.
Maybe you were born a few decades too late. I feel I was born (a few decades?) too early.
I want the world with cures for cancer, holodecks, brain uploads, immortality, jet packs, surfing on Titan, etc.
I legitimately enjoy the future we're hurtling towards. It's the most exciting thing that has ever happened to our species. I hated the world where "Apple iPhones" were the pinnacle of human achievement.
We might be able to vibe engineer our own iPhones in a few years. That'll be fun.
The fact that people who couldn't program can now build utilities and automations for themselves is beyond remarkable.
And for those of us who can engineer, we're making 10x, 20x what we were before. I'm filling my life with all kinds of fun side projects, toys, games, utility software, automations. I don't look at libraries and software packages anymore - I just build what I need.
This is like 3D printing for software.
I've built my own SRS software, my own music library, my own project management software, several agents, a video editor, a few toy video games, home automation software ... and I've done this all in a month.
I am so excited about everything I'm going to be building. Time was a limit before. There is no limit now.
> FFS we've stood on the moon.
And we slept on that for nearly 60 years. We're about to build cities there. Finally.
>The fact that people who couldn't program can now build utilities and automations for themselves is beyond remarkable.
Learning to program has never been easier or more accessible. Being able to build a website has been accessible to non-techies for years. I'll admit using an AI is convenient but "beyond remarkable" is pushing it. If anything it's a bit of a tragedy. In exchange for that convenience people are buying into a closed, third party mass surveillance infrastructure generating code they can't understand, can't repair, can't edit or redistribute. From a FOSS point of view AI is objectively worse than what we had. And that's not even getting into the flood of slop and disinformation destroying the web.
>I've built my own SRS software, my own music library, my own project management software, several agents, a video editor, a few toy video games, home automation software ... and I've done this all in a month.
I can't imagine you needed all of this done within a month and couldn't find existing solutions for most of it. It seems more like you're addicted than being productive.
And I still don't see the line between you vibe-coding a productivity app something something holograms and brain uploads and surfing on Titan.
>And we slept on that for nearly 60 years. We're about to build cities there. Finally.
No, we aren't. We're no closer to than than we were 60 years ago. If you think Elon Musk is actually going to accomplish anything of the sort I think you're bound for disappointment.
"Look daddy government, all these bad guys are doing bad things and we stopped them. You should regulate AI in the US so that companies can't use open source models or buy from China."
Ah, that makes way more sense than Anthropic's (probably deliberately misleading) insinuation that Moonshot has been burning millions of dollars in Claude API credits by swapping in a slightly better but infinitely more expensive model just to trick their users.
I get those A/B responses chatting in Gemini fairly often, and I really don't think I'd feel deceived if I later learned one of the choices was actually from a competitor's model.
I don’t think it was misleading, deliberately or otherwise. Did you read the report? I hate to call you out like that but I think you can only get that impression if you only read the above quotes. That’s not the insinuation I get at all. It’s specifically under the “illicit distillation” category. It’s never framed in anyway but as a form of distillation.
I think they are pretty fair and explicitly say “Distillation itself is a legitimate training method […] Distillation is commonly used because it reduces the resources needed to achieve more advanced capabilities”. And go on to say their definition that makes it illicit in these cases.
And, also, they almost certainly __were__ tricking users and sending their data overseas.
They mean distillation is legitimate when labs use one of their own stronger models to train a smaller one. They certainly aren’t advocating for PRC labs to distill Claude for open weight models.
FYI I've been building an app, Android first, and discovered the magic of Kotlin Multiplatform. The LLM modularized everything so that it's substantively the same code base, but with Swift only where it's needed. Still working through it, but it seems like a great and underrated platform to build on.
This is one of those “oh gosh” posts at first glance, then you actually think about it, and it’s 100% correct. Python, Node exist to save the precious commodity of developer time. Once that’s no longer precious, best value comes from purified custom Rust, no attack surface from huge libraries, optimized performance for the specific task on all possible hardware.
Windows desktop native is WinUI 3 and Windows App SDK, for better or for worse. No amount of Rust and tokens will recreate everything you get out of the box with that stack. Native needs to feel native.
The last time Windows UI felt native was maybe Windows XP. Maaayyybe 7, if we are generous. It also has to do with sluggish or no feedback by UI widgets, due to the trend of flat everywhere.
not sure why you're getting downvoted. but if you're building even very popular applications, it is quite easy to see how LLMs are very well suited for this type of consistency job.
- they follow instructions quite well.
- are tireless at doing mechanical ports between languages and frameworks.
Then why aren't people doing that? Microsoft has as much access to cheap tokens as any software company around, right, so you'd think they would be leading the way.
Keep in mind, claudecode and the other coding agents were pretty bad until around Jan of this year (2026). So it's only been about 9 months since devs have had decent coding agents and even less time has elapsed since somewhat wide adoption.
I didn't downvote but I'd expect downvotes for a completely unnuanced thought terminating cliche that ignores everything in the thread. AI fanaticism doesn't help but lots of people don't get downvoted for that alone.
At work we're moving almost everything to Rust on our backend. Massively reduced memory usage and significant latency improvements relative to our Typescript codebase. Even for code that you'd expect to work well in TS, a near 1 to 1 port to Rust has considerably improved our best case and, in particular, our worst case latencies. And the memory we get back is huge, we may even drop our instance size down with the 800MB of RAM we're likely going to save.
The type system and error handling ergonomics make it easier to write defect-free code than, say, Go or Java.
Simple servers are request scoped and mostly feature linear request handling, so you're writing simple vanilla Rust without the complex pointer semantics that you would use for systems programming. The async pieces aren't difficult either.
Serde-annotated structs are the best serialization/deserialization story anywhere. It integrates super ergonomically into Axum and Actix to make writing request handlers a breeze. They're super easy to read, too.
Compared to what? I see ASP.NET Core and Quarkus very competitive. The virtual threads in Java are great, paired with structured concurrency IMHO.
If I have to go microservice or API, I would choose FastAPI for fastest delivery and if I have to rewrite, Go, unless it is massive scale and scaling horizontally becomes a headache I would not consider Rust/C++ for this.
All the backend, etc. for the company I have been working for is C++ for the fast parts but all tools around are Python (with NiceGUI and Flask mostly).
reply