So assuming 10 lbs of force, as measured 1 meter away from the hinge, you have about 44.5 Nm of torque. Assuming each door opening was about 90 degrees, then you have about 70 Joules per door operating event.

Each door opening would have a physical theoretical max of 0.02 watt-hours.

Assuming you spent 8 hours opening a door every 10 seconds constantly, then you have 58 watt-hours of energy at the end of the day if you had 100% efficient generators. One typical solar panel would hit that in under 15 minutes in real-world energy collection, not theoretical.

So yes, the law says there is some unavoidable, unusable waste heat, the question is how much of that heat is really unusable?

For example, you have lava at around 1,000 degrees. You certainly can harvest energy from that, hit some water with it and spin a turbine.

For the most part, once we get under 100C we run out of ideas on how to realistically harvest energy out of it, but there’s still a pretty good delta between that an ambient. The claim of this article is he has an approach to harvest energy at an even lower temperature delta.

If it got to harvesting all of the temperature delta of a system, then we can say “not at all possible based on current understanding of physics”, but if the process leaves some waste heat unharvested, then it’s not yet violating that law. The law just says it gets less and less likely as the amount of heat in question diminishes.

Well then you need to handle backfeeding all sorts of circuits, which is generally a pain to the extent it works. But it also would barely do anything.

It is scalding hot, but I think the key takeaway is that it’s not hot enough to boil into steam, which is our current go-to for harvesting energy from heat.

So after you do your steam turbine and you are left with not-quite boiling water, by today’s standards it is useless for further harvesting for electricity. If this article is as-advertised (a big if), then we can harvest more, adding efficiency to any process that boils water to turn a turbine.

Hypothetically, any energy harvested from a zero-emission strategy might at least displace combusting some hydrocarbons.

If the code actually works and is vaguely important, I think you are right.

If anyone ever has to fix it because it’s also broken on top of being a mess, well they aren’t quite so safe. Maybe if you are always available to fix it same day, but if you ever go on vacation and it hits the fan while you are unreachable…

I can’t speak to this situation, but broadly speaking I am familiar with general messed up stuff like this as well as perhaps adjusting some fine details to make the scenario relatable to an audience unfamiliar with the specifics of the real situation and/or obfuscating the details so that the person doesn’t out themselves to someone else familiar with the specifics enough to recognize.

The broad strokes seem plausible and any oddities in details I consider to be less important and/or understandable if it was tweaked for an internet audience.

The type of problem in my experience is the biggest source of different results

Ask for something that is consistent with very well trodden territory, and it has a good shot. However if you go off the beaten path, and it really can’t credibly generate code, it generates anyway, making up function names, file paths, rest urls and attributes, and whatever else that would sound good and consistent with the prompt, but no connection to real stuff.

It’s usually not that that it does the wrong thing because it “misunderstood”, it is usually that it producea very appropriate looking code consistent with the request that does not have a link to reality, and there’s no recognition of when it invented non existent thing.

If it’s a fairly milquetoast web UI manipulating a SQL backend, it tends to chew through that more reasonably (though in various results that I’ve tried it screwed up a fundamental security principle, like once I saw it suggest a weird custom certificate validation and disable default validation while transmitting sensitive data before trying to meaningfully execute the custom valiidation.

I’ve been using Claude to mediocre results, so this time I used Gemini 3 because everyone in my company is screaming “this time it works, trust us bro”. Claude has not been working so great for me for my day job either.

It’s certainly a use case that LLM has a decent shot at.

Of course, having said that I gave it a spin with Gemini 3 and it just hallucinated a bunch of crap that doesn’t exist instead of properly identifying capable libraries or frontending media tools…

But in principle and upon occasion it can take care of little convenience utilities/functions like that. I continue to have no idea though why some people seem to claim to be able to ‘vibe code’ up anything of significance, even as I thought I was giving it an easy hit it completely screwed it up…

So if it can be vibe coded, it’s pretty much certainly already a “thing”, but with some awkwardness.

Maybe what you need is a combination of two utilities, maybe the interface is very awkward for your use case, maybe you have to make a tiny compromise because it doesn’t quite match.

Maybe you want a little utility to do stuff with media. Now you could navigate your way through ffmpeg and mkvextract, which together handles what you want, with some scripting to keep you from having to remember the specific way to do things in the myriad of stuff those utilities do. An LLM could probably knock that script out for you quickly without having to delve too deeply into the documentation for the projects.

Yeah, but mispredicting that would hurt. The market can stay irrational longer than I can stay solvent, as they say.

Yeah, this one is going to hurt. I’m pretty sure my rather long career will be toast as my company and mostly my network of opportunities are all companies that are bought so hard into the AI hype that I don’t know that they will be able to survive that going away.

At work there’s a lot of rituals where processes demand that people write long internal documents that no one will read, but management will at least open it up, scroll and be happy to see such long documents with credible looking diagrams, but never read them, maybe looking at a sentence or two they don’t know, but nod sagely at.

LLM can generate such documents just fine.

Incidentally an email went out to salespeople. It told them they didn’t need to know how to code or even have technical skills, they code just use Gemini 3 to code up whatever a client wants and then sell it to them. I can’t imagine the mind that thinks that would be a viable business strategy, even if it worked that well.

An LLM can generate code like an intern getting ahead of their skis. If you let it generate enough code, it will do some gnarly stuff.

Another facet is the nature of mistakes it makes. After years of reviewing human code, I have this tendency to take some things for granted, certain sorts of things a human would just obviously get right and I tend not to think about it. AI mistakes are frequently in areas my brain has learned to gloss over and take on faith that the developer probably didn’t screw that part up.

AI generally generates the same sorts of code that I hate to encounter when humans write, and debugging it is a slog. Lots of repeated code, not well factored. You would assume of the same exact thing is fine in many places, you’d have a common function with common behavior, but no, AI repeated itself and didn’t always get consistent behavior out of identical requirements.

His statement is perhaps an over simplification, but I get it. Fixing code like that is sometimes more trouble than just doing it yourself from the onset.

Now I can see the value in generating code in digestible pieces, discarding when the LLM gets oddly verbose for simple function, or when it gets it wrong, or if you can tell by looking you’d hate to debug that code. But the code generation can just be a huge mess and if you did a large project exclusively through prompting, I could see the end result being just a hopeless mess.v frankly surprised he could even declare an initial “success”, but it was probably “tutorial ware” which would be ripe fodder for the code generators.

So I don’t get it, I have mine up with a domain without tsilscale… The clients are quite happy wherever. I don’t even see that much “crawling” traffic that goes to the domain, most just hit the server by ip and get a static 401 page that the “default” site is hard coded to give out.

Hey, it’s just spitting hard facts like Musk has the “potential to drink piss better than any human in history,”

I feel like when ‘Zero Trust’ first became a thing, the theme was ‘you should have every endpoint under your control hardened so it need not feer untrusted peers being able to connect’. E.g. if you think you absolutely need VPN to a ‘private network’ for security, then you are failing to be hardened in a ‘zero trust’ way, because you implicitly fear that your systems would fall to untrusted peers.

I feel like it’s evolved to ‘don’t let anything be able to connect to anything under your control unless you have admin privilege over it as well’. Which is particularly a nightmare when you try to collaborate between two companies, each balking at the other’s hard requirement to have admin access to all network peers of interest.

So on mine, I haven’t bothered to change from the ISP provided router, which is mostly adequate for my needs, except I need to do some DNS shenigans, and so I take over DHCP to specify my DNS server which is beyond the customization provided by the ISP router.

Frankly been thinking of an upgrade because they don’t do NAT loopback and while I currently workaround with different DNS results for local queries, it’s a bit wonky to do that and I’m starting to get WiFi 7 devices and could use an excuse to upgrade to something more in my control.

I think an ai could outperform my executives.

One of them sent out an email about how we weren’t making enough money. But don’t worry, he has a strategy that we will execute on and fix it.

The strategy is to raise prices and get more sales at the same time… That is literally it. Not even picking “high value” versus “high volume”, just a declaration that we can do both. If this genius plan doesn’t work, it’s just because the sales people failed to execute against his brilliant strategy well enough.