batshit_beaver

AGI

I’m personally a huge fan of viewing LLMs as linear algebra machines. They seem a lot less fascinating when you do that.

If faster intelligence was “groundbreaking” we would still be worshipping pocket calculators.

If we’re talking about ease of entry, doesn’t entrepreneurship fit the bill? It’s not an industry per se, but I don’t think it’s ever been easier or cheaper to start a business of any kind. Succeeding is, of course, a different matter.

What do you mean there’s no frontier industry? ML, infra, robotics, and even space all fall into this category. Not to mention medicine and biotech are starting to pick up.

I’m sure out of the many attempts to automate the production of software and get rid of those pesky, expensive devs, THIS one will work.

> subscription

Well yes when your tokens are nearly free, then getting literally any business value is a good outcome.

For once I actually agree with Simon, and fortunately the percentage of engineers interested in software architecture and business concerns is so low already that I doubt they’ll have trouble in the job market. Pure “coders” were not safe from layoffs and PIPs even before LLMs, in my experience, since their jobs could so easily be outsourced and offshored.

> Software is unique in that regard that the results can be verified quickly and easily fixed

Neither of these statements are true if your team is building anything remotely complex with real customers. You can’t quickly verify if your decisions will survive the test of time and past mistakes can’t always be easily fixed. LLMs trivialize the easy parts of software development and enable faster enshittification of the projects they’re used on, but that’s about it.

You can’t abstract away the need to capture requirements in a symbolic language. It doesn’t matter which programming or formal language you choose to specify your requirements in, but it cannot be natural language.

When people install npm packages, they typically only install ones the community trusts. These packages have clear APIs, open source, good docs, test coverage, security checks, and, importantly, usually no undocumented side effects.

Engineers introduce these packages into their own projects that have far fewer of the above guarantees than OSS, are built under pressure from deadlines, and have a constantly churning set of maintainers.

I think from the perspective of a business its own code does call for greater scrutiny than its dependencies.

The first sentence is easily verifiable, but the second carries the burden of proof.

For the record I’m fully on board with slopification being a bad thing. My point is that businesses don’t seem to care at this time.

> It's your job to explain to them the consequences of shipping code faster than the people whose job it is to understand it can do so.

The problem we’re collectively running into is that nobody knows if this actually poses a problem anymore. We’ll find out in 5-10 years, but businesses and clients can’t wait that long.

As a side note, I don’t think it’s fair to blame the “management” for setting expectations around AI use. They’re driven by market forces. If competitors are producing slop and customers are gobbling it up, then from a business perspective it _is_ a mistake to abstain from AI use.

What I’m finding quite fascinating is that we’re in a weird spot right now where we’re both expected to produce slop AND understand it / be responsible for it. Seems unreasonable.

Time. The limit is time. With or without agents there’s an X number of lines of code my brain can digest per work day. But rate of production of code is effectively unlimited or heavily influenced by people who let go of at least some percentage of understanding. This forces the rest of the team to also let go to keep up. Without guardrails around the rate of code production, _teams_ tend toward 0% code understanding. Individuals working on solo projects or in tiny teams that can enforce social contracts are not affected.

Proponents of LLM coding agents would claim you don’t need reproducibility. As long as the final output passes tests/lint, it’s (allegedly) good to go.

Nobody solved the navier stokes problem, Jesus. A subproblem was possibly solved (still being verified) based on context in an actual mathematician’s chats with ChatGPT. These models on their own are still incapable of producing anything but slop.

Absolute cinema - measuring software engineering productivity by lines of code in 2026.

Trying to turn it around by once again claiming that ai is an existential threat and will take everyone’s job right before brutally murdering every human.

These companies don’t appear to give a single fuck about public perception. Only their investors and the government (in anticipation of the biggest bailout in human history).

> I've found I can produce 10x more code than I could otherwise, of lower quality than I would otherwise, but the speedup is worth it.

The speedup of slop production being “worth it” is what we, as a society, are having trouble evaluating at this point in time. In all likelihood it’s worth it only in the short term.

There is, however, a question of scale.

And yet those companies aren't collapsing because it became easier to rewrite their software, but because llms themselves have replaced a core part of their business (such as translation or simple art).

Seems like so far the market is not convinced at all that SaaS companies as a whole are any more disruptable than before.

It’s not just math. Anecdotally, LLMs struggle the same way with software engineering where the code they write is correct (compiles and passes tests), but reasoning is wrong often enough to eliminate most trust in these models’ ability to explain codebases or even features they themselves produce. It’s not about the _level_ of the supposed intelligence where a model struggles to summarize things succinctly or simply enough (responding to the “pocket mathematician” comment) or can’t grasp certain concepts at all (if so, how tf is it able to apply them?). It’s that by their design LLMs have no concept of truth and no concept of causality. They guess with every single inference and it’s very hard as a user to understand which guesses are more or less certain, since, you know, confidence ratings aren’t part of these models’ design either.

They score 99% but then can’t explain _why_ their solution works with any accuracy. Not as useful for learning as one might think.

It was this way well before vibe coding. Over a decade of zero interest rates combined with talent wars and other anticompetitive behaviors by large tech companies did the industry in.

> Now back to AI for software creation. "Review is the bottleneck because EVERYTHING must be judged by a human." No. It doesn't. We just need to build AI review systems, that will do reviews better than (or at least as good as) humans. The human review quality bar is far below 100%. Far far far. If we can show (likely in the next 12-24 months I think) that AI review quality is consistently above the human review quality - again, based on evidence, based on real data - then that's it, then there's no good reason to have humans review the code.

Let me know once the majority of software engineering organizations start only checking in markdown files and let code be generated non-deterministically from these specs in CI. If this is not happening now, there’s clearly a sufficiently high level of distrust in blind LLM output (both code itself and reviews).

And even then, are you suggesting humans shouldn’t at least review the markdown specs? Why not have LLMs review the specs then? Is there, perhaps, some fundamental quality to human review process that is desirable here?

The challenge with comparing these things to humans, is that humans learn. A newbie might not respect your organization’s set of policies on day one, but what about 3 months in? Or 3 years? Meanwhile there’s still no reasonable mechanism for automatically fine tuning LLMs or adjusting their harnesses to make them better at completing your organization’s objectives more successfully. They’re still overwhelmingly governed by the shared weights and harness policies found to be successful for the average case.

I think the issue is that obtaining empirical proof of AI or manual coding being more efficient is very difficult, since true costs and outcomes aren’t known for months and often years. AI _is_ faster at producing short term results though, so naturally, given the state of the industry, everyone is piling both money and time into AI driven workflows and manual work is met with suspicion, if not outright discouragement. Engineers might hate AI, but they need their paychecks.

Hopefully everyone? Else your job could have been outsourced or replaced by a junior with access to Google and StackOverflow way before LLMs (it just wasn’t due to zero interest rates and proliferation of bullshit jobs in tech companies).

Humans have goal seeking behavior. LLMs don’t. You could maybe call the combination of LLMs and the RL-based harnesses somewhat “intelligent” in aggregate, but the problem is that it’s not “general” intelligence like these labs want to argue, since it’s by definition only good for the set of problems the RL part has been trained to solve, which is a subset of programming problems.