I thought a big part of the reason for type systems was a sort of self documentation/contract? Especially if you need to work on an unfamiliar system with bad documentation. Also what about system boundaries? I prefer typed languages personally.
The benefit is not only about "documenting the contracts" but documenting the contracts in a way that we can trust those contracts can not be violated when the program is running.
That is a very good thing to help us reason about the program, we have invariants we know must hold true if the program does not stop in a type-error.
If you asked me last year(2025) I would have still said LLMs are a silly toy.
As of Jan 2026 I have come to accept that LLMs are at least part of the puzzle of how intelligence works. They are at this point better than the majority of humans at various intellectual tasks. It may not be or ever be a 1:1 but good enough ran the world already before llms.
There is not even a formal definition of what intelligence is so saying LLM's are intelligent can't even be "right/wrong". Its just arguing semantics and definitions.
They are better than humans at tasks that require information recall and application to specific task.
For example, front end web app layout and basic functionality. Anyone can make a website with interactive buttons with ease now, where as before, you had to go look up examples, try stuff, figure out why its not working, e.t.c.
But in terms of organization and higher level tasks, like for example making front end that is clean, robust, easily extensible, and doesn't break, LLMs require almost as much prompting to do this as it takes to actually write the code.
They are mostly "faster" than the majority of humans. They are rarely better than experienced and talented humans at the majority of tasks they are able to do. They are better on both scales on a small thin slice of work tasks.
They're not better than the best humans at practically anything. However I doubt there's a person alive that could outperform an LLM on a broad suite of tasks like Humanity's Last Exam and the vast majority of people probably couldn't answer a single question on it.
They’re the language part of the puzzle, which seems to require some basic world modeling but it can’t make novel models unless there’s an example in its training data.
I think engineering and mathematical thought requires spatial reasoning, when I model problems I see them as 3D shapes. Like the economy is a series of tubes that money flows through and collect in buckets, programming state is little boxes that hold values, chemical interactions are like keys that fit into locks.
I don’t think LLMs can build models like that, but because it has so much memorized and there usually isn’t a need for a novel model custom fit for a problem, it can fake it by imitation.
Seems like LLMs are that. A bunch of most probable word associations is a network, and you can build a physical model of a network, or build a network that allows you to reason about a physical model. Whether it's just a flowchart or workflow diagram, or an X-dimensional matrix with vectors moving through it.
But the only way to map the network in an LLM is experimentally. You have to prompt it, and see how the coefficients fall in order to construct your most likely walk through the training data.
I think that LLMs can and do come up with novel things through exhaustion, just by applying the relationships between some set of entities to entirely different sets of entities because an accumulation of earlier context pushed the probability of those entities being mentioned, and they were able to easily replace a selection of entities that were more associated with those nearer connective, relationship words.
I think that as such LLMs are good at generating metaphors, and a lot of innovation comes from going "What if As worked like Bs?" Just go through all the As and Bs, toss the ones that don't make any sense and test the ones that seem like they might.
I don't believe you can say that "LLM" is part of intelligence. No single human is exposed to as much text as any LLM model ingests, not by many orders of magnitude, and humans still perform cognition and generate new language.
Most LLMs are multimodal now, able to map visual concepts to language and vice versa. If OpenAI's recent Erdos solution was faking math, it faked it very well.
3D isn’t one of the modes though, I know a paper several years back showed that diffusion models don’t actually understand physics or geometry.
I can’t evaluate the Erdos solution personally, but both math and software have many problems that are some combination of other problems and since it can get instant verification feedback it can try millions of permutations to discover the right solution. This is valuable, I’m not dismissing it, but I think there’s another tier of harder problems that I don’t believe LLMs can solve and it will require some further theoretical breakthroughs to get there.
"How can we understand what an LLM is "thinking"? It's clearly very valuable to do so — it could enable steering model behavior, detecting dangerous intent, and more."
Well that is complete any utter bollocks, dribbled in para three or so, and obviously written by a next token guesser.
LLMs are tools and I'm pretty sure if I let you loose on some of my tools, you might lose an extremity unless I kept an eye on you.
I have an on prem Qwen3.6-35B-A3B-UD-Q4_K_XL working on a box in the office and its quite handy for a chat.
I moved to fastmail about 2 years ago. It was not a super painful process, they make it easy as possible. Also I used it as an opportunity to "reverse-whitelist" all my accounts by giving everything i use its own disposable masked address.
If you want human connection the legal system is not where you are going to find it, period.
I don't think there will be any such market for "non ai" law. If I'm involved with the legal system I just want out as quick as possible as cheap as possible.
Bad legal advice will keep you dealing with the legal system for much longer and at much greater cost. Something being cheap and quick upfront doesn't mean it will be cheap and quick by the end of the process.
Maybe, although I would be extremely hesitant to extrapolate from this one study and trust my legal life to an LLM. One thing that's worth noting, though, is that regardless of the quality of objective legal advice in the abstract, for a lot of smaller scale stuff the human connection actually is literally what is important. There are ambiguities in the law, which are not resolved deterministically but rather at the individual discretion of judges. Your lawyer, if they're any good at their job, knows the local judges and how they're likely to rule for given circumstances, which can influence their legal advice to you specifically.
The legals system is structurally based around manipulating text and its relations. It seems to me that the entire legal industry is the ideal use case for LLM's to take over.
Of course the legal system can gatekeep forever by design.
For Decades the huge tech companies basically faced no adversity whatsoever. Now for the first time in their existence the massive returned investments in AI they are experiencing ... we will call it pain.
I would say it will be interesting to see what they do but I think rent-seeking, oppression, human rights violations would be more apt.
They were of course trustworthy proviers while they were untouchable but now I know how things are gonna go.
The main reasons those places lost support is they became convenient prisons without due process. Why do you think there are so many horror movies based on the setting of a sane person involuntarily put there?
While not ideal you gotta admit now that those people that need help are in your face rather than conveniently disappeared you are thinking about their plight some.
Maybe try to think of something better than forever prisons and stop becoming a ghoul.
My problem with most of these books is they are indirectly trying to solve the real problem. The problem that IME HN is allergic to discussing.
Power Dynamics.
The reason the CEO is nitpicking your job is because he is not a good CEO and doesn't know his place or how to do his job. Almost all these books are about an indirect way of dealing with the fact that, this person is a ID10T and you have to deal with them because they have more power than you. Yet it is literally NEVER discussed.
The books(IDK about this one) really summarizes indirect ways of how to be subservient and not accidentally antagonize your "superiors" which are frequently people just born into a better lot in life than you, without feeling like that is what you are doing.
What is the CEO's primary duties, networking?, Sales, COMMUNICATING yet its your job to read books on how to tiptoe around how to sus out what they cannot COMMUNICATE?
I'm a pretty opinionated engineer but I'll still volunteer that in a majority of "engineering" disputes, I care more about having a coordinated and consistent approach than I do about the absolute tack taken.
Maybe I've just been lucky to mostly work with decent managers, but basically I consider the tie-breaking function to be intrinsically valuable.
With this particular book, the prerequisite is that your client is trying to achieve something, yeah. I think know the type of CEOs and CTOs that you’re talking about, the ones that only want to sound smart and don’t really care about the end result. Unfortunately, there’s not much you can do in this case apart from looking for a workplace where people do care about what they do.
We do it like that with everything. If you consider yourself an artist it is quite simple to say you cant put your name on it if anything changes. You can also explain what you just wrote: Youve hired me, trust me to do it and focus on your tasks. Or: we will be different from otters but in a limited number of ways and your suggestions dont offer enough roi to make the cut.
I know. And it’s what I learned from Jack Palace on the Ripley’s Believe It Or Not TV show, where I first heard about it. Because it makes a really good story.
“She was a woman so sexism meant couldn’t be an architect. The only way she could live out her dream was just to build her own house over and over.“ isn’t nearly as fun.
The fact she was afraid of ghosts from people killed by her husband‘s invention just slots in so well as “evidence“. Add a crazy looking house and there you go, perfect tourist trap.
>DevOps won't want to support it? Burn down the company
Still true but you seem to blame the ops. I've been in a job where every dept was allowed free for all tech budgets. They would hire incompetent consultants to dump 3000hrs of work on devops then do it again next week and complain about how devops never gets anything done. Then 5 other departments would do the same thing.
You know how 99% of the work is in that last 5% of the project. Thats how all those consultants would leave everything.
I read that as a frustration with the disparity between "you build it, you run it" and the enterprise-y habit to co-opt terms from free-roaming developers and stripping them of all meaning.
You can still have a central team of operators. When they're expected to deploy and support applications from development or procurement teams, I'd argue that's something else than devops for better or worse.
what I meant with "DevOps won't want to support it" was someone saying this before DevOps had even been asked, and by someone who wasn't even on DevOps, who just assumed that they probably wouldn't like this sort of thing.
If this meeting doesn't include devs, I don't know what the context of anything in the article is supposed to be about. There are no major new ideas in business that don't involve software.
If it is not the case that "devs" is not functionally equivalent to saying "DevOps", then "DevOps" doesn't exist. You have an operations group, and you need their buy-in, so they should be invited to the meeting.
reply