One of the big problems with this is that it effectively changes the job description of programmers to managers.

You're not just responsible for your own code, you're responsible for telling someone/something else what to do, and then reviewing their work, and then integrating that work and making sure it's correct, because if it gets checked in and it's wrong, you're on the hook for it. But you didn't even write it, so it's harder to understand. Most programmers I know are better at understanding the code they wrote rather than the code they read.

Management is not a job for everyone. I've been a senior programmer and a lead programmer, and it is easier to be a senior programmer that writes complicated, essential systems than it is to be a lead programmer that REVIEWS those systems. If you don't want to do that kind of management work and you just enjoy writing code, you should stay a senior programmer, or move up to some other non-management job. But if you're forced to babysit an LLM, you get ALL THE WORST PARTS of being a lead programmer. You only get the code reviews and none of the intellectual stimulation and enjoyment of creating that thing on your own.

At least a couple comments have pointed out that solo devs use LLMs to great effect, and I'm sure this is true. But if you're not a solo dev, the task is so much different. You're trying to integrate with a bunch of other people and other systems and you're on the hook for something you didn't even write, so you'd best spend the time learning what the code says. It completely defeats the point. When I get an LLM to write my powershell scripts for me, it's because I have no interest at all in learning that scripting language. It takes 2 minutes, usually does a good job, and I can forget about it forever. The development model of asking an LLM to write code that you're responsible for that has to intermingle with the code of dozens of other people and agents sounds like a straight-up nightmare.

Solo development is a different job. The tasks you give to AI are based on what you need, and your own mental model of they code they need to spit out.

Large-scale development is different, because people are different. I've worked on 5 or 6 different game engines over the years, and Unreal is the current one that I'm on. Unreal has its quirks. The rest of our game engine has historical baggage that I need to account for. The interactions between systems are more organic, because not everything can be planned in advance. I worked on Assassin's Creed for 3 games (about 7-8 years?) and we carry over the mistakes of the previous games to the next one, because that engine shipped a game, so it's a known quantity. Sometimes if you're lucky, you get to fix some of those bugs, but then the next games get YOUR bugs, etc.

Solo devs ARE doing amazing work with AI, but solo devs *by definition* don't need to understand the code of 1 or 4 or 50 other people. You can't just check shit in on a big team.

Perfectly reasonable study, because small-scale development is not the same. I have to understand the code of 10+ programmers on a regular basis. Our game engine is involved. The only way for me to learn the code is to use it and step through it.

If you're writing apps yourself, AI is great. It's a teammate that follows your instructions and implements the things that you want. You've already got a mental model of what you need, and it gives that to you. Of course it works for you.

Big codebases are just different. Systems are interdepedent in weird ways for historical reasons, or vendor reasons or weird coworker reasons or whatever. It's not the same task.

Yeah, PC sales were artificially high in 2025 because of the Windows 11 requirements, now prices have rocketed so sales are artificially low. The next step is presumably when the AI companies - and their investors - realise that they are not going to make a profit on all those tokens and some go belly-up. The Chinese are also going to be making more RAM and SSDs for their own market to avoid sanctions, prices are going to have to drop.

(from 1987) has a section on how excessive use of nitrogenous fertilizer (as in most conventional agriculture) changes the electrostatic potential on clay micelles via pH changes which caused many of the other nutrients in the soil which are no longer as strongly held by clay to leach out (related to changing and so become unavailable to plants:
https://archive.org/details/to...

Having more organic matter in the soil can help resists this somewhat, but the real solution is not to dump too much chemical fertilizer on the soil and to consider using rock dust and other amendments to slowly remineralize the soil.
https://remineralize.org/

So, an aspect of this may be that many farmers may think more is better with chemical fertilizer (even as depleted soils, problematical cultivars, and early harvesting may indeed be aspects too).

As above, this has all been known for *decades* but it is hard to get a lot of people to care about it because the health effects are so gradual and the short-term profits from conventional agriculture are so immediate.

Good advice as always though to eat a varied balanced diet -- including foods grown in various places.

>Please just put a sock in it.

No.

The OP complained about low quality programmers and "slop"; clearly a reference to AI generated code.

My retort was to challenge that - to point out that fronteir LLMs are now much better at writing software than most programmers. That is objectively true. It doesn't "feel true"; it _is_ true.

Then someone replied to me to try and snipe at ESR and DHH, as if they weren't both excellent programmers. Both of them are. That wasn't a person arguing against my point; that was the response of someone who COULDNT argue against my point, but was upset about it.

There are some bright and accomplished people here. That's part of what makes the place interesting and valuable.

There are also a lot of drive by haters.

Those are two examples.

You can also go read what Carmack says about it. (example: x.com post from 2026/9/11)

I hope you won't be so stupid so as to try dunking on Carmack.

I apologize for not knowing you or your accomplishments, but just playing the law of averages, both ESR and DHH have significantly larger bodies of successfully and daily-used code than you do, with bigger impacts on the field and the industry, than you do.

You probably don't have the right to shit on them, given your own lack of contributions.

(But to be fair, I'm just assuming, based on the law of averages. I apologize if I should know you and your great contributions, but do not)

I can say that both of them certainly eclipse _me_ so greatly that I can't be seen on any map that is zoomed out enough to show both of them. And while I am a relative nothing, I've had a handful of patches accepted in my life... and I've secured a few software patents in my day. That puts me in above-average territory. But there's more accomplishment gap between them and I, than there is between me and a person who cannot code at all.

There's no harm in admiring outstanding people. A particular fault of our age is that men do not know how to truly compliment and admire greatness when they find it.

ESR is a great man. DHH is a great man. My world and our industry is better for both of them. When they say that LLMs have completely changed the way they do the thing that they are better than almost all other human beings at... you should listen.

I agree we need better programmers.

We have better programmers.

The frontier models _are_ better programmers than all but the very best.

Look at the AI related posts from ESR and DHH on twitter. They are _all in_ on AI.

(DHH has been showing benchmarks of web apps being machine-re-written in Rust. Unsurprisingly, the apps get much faster)

A rough estimate is that the best LLMs do software engineering about as well as IQ 150 software engineers do it. That makes them better at it than the vast majority of people who write code.

People see an AI do something they dont like and then conclude "the AI sucks". But lets use honest weights and measures: Human programmers make worse mistakes, and make them more often, and iterate 1000x more slowly.

The way to think about it is this: generally the AI will write better code than the person prompting it would write, and it will do it several OOM faster. That means more code getting written than would otherwise get written, AND, it means that code is higher quality than it would be otherwise.

"Perfect" isn't the bar. "Better and in less time than if the human doing the prompting had written it themselves" is the bar. And that bar has been cleared with by a mile of daylight.

Junior engineers don't normally get to tell IQ 150 engineers what to go do. LLMs change that.

There's quite a lot of nuance here, I guess, which isn't unexpected. It's quite a useful anecdote. At the very least, we can't just wave it away like it's not useful. It may generate sub-standard paintings or completely miss the point of art, but if it makes your proofs better, it seems insane to not have it help.

Sure.

That was not the case here.

Microsoft did not require masters degrees for entry level dev jobs.

It's common for Indian/Chinese folks to _have_ masters degrees, because going to the US for graduate school is a common "first step" in the process to get their feet on US soil - F1 student visas.

The way it typically works for an Indian or Chinese software engineer to get into Microsoft (or any of the majors, I assume)

- get 4 year degree (real or fake) back home
- get accepted to US grad school; get student VISA (F1)
- do grad school
- look for US company that will sponsor/supply an H1-B slot
- get OPT status for STEM (up to 3 yr extension of residence permit w/ industry work authorization)
- start at new company under H1-B status
- start the Greencard process

This is the normal path.

Now, here is what is structurally fraudulent about this.

1) The F1 visa is intended for students who plan to leave when they are done. It is not intended to be the on-ramp into the US labor force.
2) the OPT extension to the F1 VISA is not intended to be a long term resident work permit. They max out at 3 years for software people
3) Greencard sponsorship in this scenario is supposed to be downstream of the assertion that there is no human being already in America willing/able to do the work.

From a company POV
The converion rate from OPT to H1B varies; the # of H1B lottery slots are fixed, and people who run out their OPT clock without securing an H1-B lottery slot are allegedly supposed to go home.

The corps know that. So the OPT period essentially turns out to be a "no risk try out" where the company gets to see who is productive and supplicant.

This continues once someone has an H1B; that person knows that the moment they lose their job, they have n days to find a different H1B slot before they are subject to a removal order.

An H1B friend explained it to me: as long as he's on H1B, he never says no to working late; he never says no to taking the extra incident case, he never says no.

So he wants a Greencard for safety/security reasons; so he can stop being as much of an indentured servant.

The colleges love this system because F1 students can be charged full price to attend. The companies love this because they get indentured servants with a tryout period. And because they have an infinite supply of young 3rd world labor who is fighting over the scraps of available slots, which causes dramatic wage suppression.

When they find a gem, they can move forward with a fraudulent PERM claim and start the greencard sponsorhsip process.

There may be other things going on here. If you've been in the industry long enough, you may have come across the claim that "indians hire indians", and then if you look back at all the teams you've been on, the claim appears to fit the observed data.

(which is not to say that the claim is _true_... just that my experiences don't refute it)

the human population will have expanded so far as to approach infinity and thus will automatically annihilate itself

Nope - wrong reason: it's actually going to happen because Trump will bring about chaos at the midterms, causing civil unrest in the USA, which he'll then try to distract from by dropping a nuke somewhere.

> unless said workers have the same skills and experience.

At Microsoft, its usually the hypocritical scenario.

There were very few specialized dev roles at Microsoft. Either you're thought to be smart enough and adaptable enough to work on whatever you're told to do, or, you aren't hired.

Exceptions would be "face characters" - unique individuals with unique contributions; people slashdot might have heard of. But those are far less than 1% of the company.

When you are cutting staff by thousands and hiring staff by thousands, none of them are special, none of them are differentiated. Its just bodies.

The layoff policy at Microsoft when I was observing it was pretty brutal. Each round was slightly different, but the mechanisms were usually similar: Each layer of management hierarchy, starting at the top, would get an "NTE" number, "Not to Exceed". As in, "your headcount is NTE 514" or whatever the number was for that manager. And whatever number you got, you subdivided it amongst the mangers beneath you, they subdivided to their child managers, and on downards, so most people in certain org-trees had some quota of people they were responsible for cutting.

Often, the layoffs were not based on the skills the person had, or how good or bad they were at their job, or anything like that. Usually entire teams would be eliminated, even if they had great people the company should have wanted to keep.

Yes, you did sometimes have awful situations where you lost really outstanding people just because they were on the wrong team when it came time to conduct random firings/layoffs.

The company seemed to have the idea that if they were good, they would reapply and find their way back in at some point.

The mechanism had a certain sort of "trapped by corporate America stupidity" feel to it. "We need to cut xx people and we can't be very intelligent or selective about it.. we cant make it about the specific person [because that invites lawsuits], so we make it about 'reducing investment in product [xyz]"

I am aware of possibly 2-3 RIF ("reduction in force") actions where the performance of the affected individuals was a factor in them getting chosen. More often it was whole-team cuts.

I was a dev lead at Microsoft. I believe I was asked to support immigration fraud.

I had an employee on a VISA. He was in a junior level role.

The immigration attorney asked me to verify that the job requirements for his job were:

- masters degree
- expertise in python
- (a few other things)

He had a masters degree. He had the other things on the list.

Problem was, the job he actually had was an entry level dev job. It was the exact job I got hired in at at the beginning of my career. For the level/job he was at, a 4 year degree is normal. Python (or any other specific language) is not needed.

So i said "no, that's not right. You don't need this, or this, or that"

And the immigration attorney was like "OH! My mistake! It sounds like this is more of an entry level position! Well, I guess based on what you've told me, we cannot sponsor the Greencard process for this employee at this time"

I had no idea that's what was going on or what was at stake when I answered the question. My employee wasnt mad at me - he quit MS and transferred to Google - who promised to push through his Greencard process. He said once he got done at Google he looked forward to working for me again at Microsoft some day.

So you tell me what you think was going on.

I think I was supposed to rubberstamp Microsoft immigration fraud, and I accidentally didn't, because I tried to accurately answer the question I was actually asked.

But look at how the lawyer responded. "Oh, shoot, what a silly mistake - nevermind!".

So no matter how I responded, he has plausible deniability. And if I had said, "yes, these are the required skills", nobody would have known otherwise.

I was gutted that I had ruined things for my employee without realizing it. The fact that he wasn't furious with me surprised me. To him, it was all a transaction. He wanted a Greencard; so he was going to go where he could get it - no hard feelings.

It's all transactional.

>OK, you don't work in software. Good tech companies don't outsource core development. They hire in house.

I was a software engineer at Microsoft for 22 years.

For all 22 of those years, Microsoft had core developers on every product who were not Microsoft employees.

You may want to look up the "permatemps" lawsuit from the early 00s. Microsoft had so many permanant contractor "non employees" that they were succesfully sued for the practice.

Microsoft has all kinds of legal sub-structures and "at a distance" legal entities for various employment arbitrage reasons. The technical talent of the people is not a determining factor in whether someone is a contractor, vendor, or full time employee.... Or if they are employed at an overseas subsidiary or domestically here in the USA.

And Microsoft plays a LOT of games with hiring/employment laws.

For instance, Microsoft had (has?) a large Vancouver development shop because, as I understand it:
1) its faster to get Chinese/Indian nationals into Canada than it is to get them into the USA
2) Vancouver is just up the road from Redmond, so easy to come down to Redmond for in-person meetings once in a while
3) It's very fast, legally to get people from Canada to the USA

So It was an open secret that MS would hire foreigners into the Vancouver office and park them there for, IIRC, 1 year, before moving them into the US.

Well, wait, I don't think that actually disproves the claim. I'm not saying it's not impressive, I just don't think it's as novel as you say.

When I ask an LLM to write me code that does something 'novel', it usually does quite a good job. I'm just using the freely available models, but they quite competently write code that is not truly derivative of any other code. If I give them some code of my own that has a bug in it, they'll find the bug and patch it, and that bug is, for all intents and purposes, unique.

You're still seeing a bunch of old patterns rehashed--I presume that other proofs introduce infinity families, and they can iterate through inside a clear logical structure to test whether or not something works. Again, this is impressive work, but I'm not personally convinced that this is 'novel' in terms of the broader corpus of mathematical knowledge. It didn't invent a new field of mathematics to solve your problem.

Even if you didn't prompt it, take a look sometimes at the contemplations of the models, if that's available to you. Before the model even arrives in your hands, it's got stacks of prompts giving it meta instructions, and it then generates prompts for itself to check your work. I think it would be interesting if you ran your experiment again on a new account that's fresh and see if you can watch the chain of reasoning. You might be right, it MAY be doing something TRULY novel, but as I've said, I've got my doubts.

Obviously, you're a mathematician and I'm not; I can't tell you that you're wrong about it if you insist that it is doing something new.