CyLith

One can argue that living apart from society is aiming to fulfill the same goals as those trying to colonize Mars; it is an effort to shore up human resiliency in the face of an uncertain and possibly less hospitable future of the planet. The skills required to bootstrap civilization and the technological stack need to be practiced and exercised so they do not die out, and this kind of off-grid living is very much keeping those threads alive.

Well, if you really believe that, then we truly are screwed. Trusting humans to not mis-use something is wishful thinking.

I see all these comments here dunking on the article, but I have to say, I basically completely agree with all the points the author is making. I think there are different kinds of software engineers, and the author's point of view resonates with those of us that are mostly concerned with building tools, rather than products. At the end of the day, I need the tool to work so that it can be used to build other things. In that sense, coding is dead for me. Anything more than trivial few dozen line changes to the codebase is going to be done by claude. It will figure out the hidden dependencies and gotchas more carefully than I will on a given time budget. And for me, unit tests have always been an afterthought, so the fact that the agent will also add unit tests for its modifications means it's strictly more disciplined than I ever was. I've been at this for 25+ years, and had the fortune to be able to largely be a "cowboy coder" but those days are long gone.

I still love writing code myself, but as I use LLMs more, I'm realizing it's not really the coding I actually loved, it was the software architecture and up front design work. Once the structure and plans are laid out, writing the code is just mechanical, and I'm going to reach for an agent if I can. Of course, these days, I find myself asking claude for advice on architecture as well, but that is much less a deferral of responsibilities than asking for an audit or survey.

I use Claude Opus 4.8 almost exclusively. I have had fairly good experiences with it. One time it derived an entirely novel simulation method different than anything in literature by combining its knowledge about how problems in other fields with similar underlying mathematical structure are solved. That was a bit of a Jacobian Conjecture moment for me.

I think I'm probably in the minority here, but for my line of work, the software engineering and coding is only a small part of the work. I write simulation software, so a deep understanding of physics, math, and how they can be applied to the software is absolutely crucial. I'm assuming this model is tuned to be more focused on SWE topics, and the very reason we seek "multidisciplinary" hires is the also why I actually need a jack-of-all-trades model to back my coding agents.

I'm having it help design a 68k computer similar in spirit to the original Mac (the spirit being a tightly coupled video subsystem that time-shares the CPU bus), but updated with more modern peripherals, like PS/2 and SD cards. It's got the design more or less done, but the routing will be a nightmare. I'm not ready to just gamble on it having gotten everything right, so I will be doing a thorough design review myself and re-deriving all the timing analysis.

Because living longer is a huge drain on resources that could be better spent on other things. End of life care is expensive and rarely results in a "good" life for the the life being extended.

A personal data point: I had Claude Opus 4.8 design a fairly textbook circuit that outputs a monochrome image burned in an EEPROM over standard 640x480 VGA using only 74 series logic and GALs. It designed the circuit and GAL code, and I did the routing, and got it made through JLC for $6. After it came back, there was one error that was not caught, which I could blue-wire, and it works just fine otherwise. I was fairly impressed.

This was the single worst thing about working at Google. Every PhD thinks they're smarter than having to do the actual work. Sometimes you just have to do the fucking work.

I think what people are taking issue with is that the machine is not accessible to everyone; OpenAI and Anthropic have monetary incentive to not only gatekeep the knowledge acquired, but also to destroy the original copies, or make them impossibly difficult to find.

That’s funny, when I first learned about the equivalence principle, my first thought was “of course!” I have always found it to be very intuitive. The great leap is being able to frame it that way.

I know relatively little about the workings of LLMs, but I keep seeing projects like this that run massive MoE models using very modest amounts of RAM, perhaps excessively so. I wonder, is there a way to make the RAM usage tunable? I have a Macbook with 32 GB of RAM, and it'd be great if I could run the same model but take advantage of the additional RAM to make it run faster.

The issue I have with all these formally verified numerical algorithms is that, at the end of the day, I need this implemented using actual floating point. It's great that it's verified, but not for the case that is practically meaningful. Unless it's implemented using hardware accelerated floating point, most of these algorithms are not sufficiently performant to be practically useful.

I feel like if the logic in git were just a bit more sophisticated, many of these issues wouldn't even arise. Most of these are because someone foolishly (usually me) merged in the wrong direction for convenience's sake, and now I have a duplicated set of commits. It's not like Claude is doing anything very sophisticated; it's just looking at commit hashes and seeing what the unique changes are.

These days, I am leaning heavily on Claude to deal with all the git issues that I never wanted to be bothered with. A rebase that can't just fast-forward? That's a job for Claude. A merge conflict that really shouldn't be a problem? Claude deals with it. These are things I never wanted to have to learn in the first place, and I don't feel like reserving a part of my brain to remember how to deal with these things is worthwhile. In almost all cases, there is no actual decision that needs to be made, it's purely a mechanical process.