jaggederest

It's funny because that strikes me as a perfectly fine thing to do for a Brian Eno docu, but for virtually every other program on the planet as absolutely egregious. I guess there's always an exception!

Could go even further and produce multiple sections and let the "algorithm" edit them together in whatever order.

intro (2x 15 second clips pulled from random places in all the other sections)

section 1 (a/b/c)

section 2 (a/b/c/omit)

section 3 (short/long)

section 4 (a/b/c)

It would be like a choose your own adventure video, without the choosing, or the adventure.

I suspect this is going to end up like most services provided e.g. cloud stuff, balkanized between a couple major players and an assortment of DIY or less popular options if you don't like those ecosystems, plus some UX/DX focused wrappers that use the big players under the hood.

I think that would be a pretty satisfactory outcome compared to one hypercompany consuming trillions of dollars of the world economy.

Absolutely can pull the water you need for e.g. household purposes out of the air - somewhere north of $0.10 a gallon or so, spendy, but close to trucked water - maybe $1k a month, give or take. People pay that much for trucked water, though a well is rapidly cheaper if it's available.

For a farm, say growing almonds (notably water intensive, but a reasonable first pass), you need about 4 acre-feet per acre per year, which would cost... hmm, somewhere in the very rough ballpark of $200k / yr.

That might put the economics of almond farming substantially in doubt, given that an acre of almonds might yield a value (revenue, not profit) of, say, $10k on a good year.

The thing I like to do is to use models from different training sets - so for frontier, OpenAI criticizes Anthropic and vice versa. They're very much peanut butter and chocolate in that regard - I honestly can't be bothered to set up the whole MMLQUALA benchmark suites or anything, but I wonder how high "the two best models running at max thinking working together" would score compared to either individually.

Similarly to your, ahem, long-orbit satellite issue, og Iridium flares were/are so bright they made/make the moon look dim by comparison.

The issue is that so much of night time lighting is wasteful and excessive, not that lighting doesn't enhance the world for humans.

I know for a fact that just putting old school light shutters on streetlights so they don't leak into, say, bedrooms and stay focused on pavement and sidewalks, and dimming them off hours, would be a huge change that doesn't diminish utility at all.

Huckleberries are also lifechanging. There are stories that people protect their favorite patches with force of arms.

Funnily enough they aren't really commercially cultivated because they need native habitat, as far as I can tell, but my parents planted a bunch as landscaping and they're happily churning out berries 50 years later.

Wait till you see what's coming up soon. The commercial nurseries are creating some really insane stonefruit and other things like that, especially Floyd Zaiger, but they haven't become widely planted or popular yet, and stonefruit take several years to get established, so expect another profusion of fruit varieties over the next decade or so.

Flavor King is an older varietal (1991) you might hear about. Apparently they can even cause a theft problem when planted in urban areas along fences, because people will borrow a pluot, try it, and come back with a wheelbarrow they're so tasty.

How certain are you about that? https://guides.rubyonrails.org/v2.3/plugins.html

Back in the very early days you would have used script/plugin install (svn url)

Here's one of my mirror repositories migrated from SVN into git. https://github.com/jaggederest/acts_as_voteable

Other animals do get heart cancer, which is fascinating. It's still very uncommon compared to other cancers, but not virtually absent as it is in people.

I think the issue with Rails, as a long time developer in that ecosystem, is that much of the default assumptions are a) not common (not so far off-piste that a human can't get used to them, but outside the center of the bell curve), and b) not explicit. So it's very common to do things like install a plugin that has across-the-repo effects but that is only documented in e.g. the config directory and the bundle. Unlike, say, NPM, when you install a new package where you explicitly have to e.g. import it everywhere you want to use it.

As time has gone on Rails has gotten better about this, but it still means that "your rails" and "my rails" are different enough that AI could easily get confused.

As you say, the other aspect is the intense dynamism which works against AI, which again prefers to have everything explicit and reasonable about from the text.

So I ended up going back to Javascript after a brief 12 year digression into Ruby. I still love Ruby but it's just not the right language for the current environment and team structures - typescript is "worse is better" in the right ways that it's a much better choice for both frontend/backend being the same and better type systems.

We looked at the Solterra in 2025 and it was genuinely awful, I can't believe anyone who works at a car company looked at it and thought "this is substantially better and will sell more than the Pontiac Aztek despite aesthetically and practically being worse in many regards". The interior center console feels like a prank, almost.

Toyota, like many companies that succeed based on merits beyond the aesthetic, secretly dreams of being an aesthetics-first company like, say, Porche or Ferrari, and always tries their hardest to reject reality so that they can do the sexy-but-awful thing instead of the amazing-but-ugly thing.

You see this recently over the last couple years in tech where everyone is "an AI company" even when they make something totally unrelated - same impulse, different industry.

Bro it's so easy just save 20% of what you need above and beyond the basics of an upper middle class living, like maybe take one fewer trip to Bali each year, or cut back to only two cars per person, or timeshare your pool.

I first heard of it here: https://ludens.cl/paradise/turbine/turbine.html (CTRL+F transformer), a pretty fascinating DIY medium distance power transmission setup.

Would have been very interested. I know there are genius/insane people out there winding their own transformers.

dd if=/dev/urandom of=/some/website

Seems deece

Correct. Delta t

There needs to be a gradient or the wax won't move due to temperature. The motion is the change in state, as you say, but it won't change state if you don't heat or cool it.

That's why it's a motor, like in dishwashers, not a thermostat like in e.g. greenhouse windows or radiators.

It doesn't need to be persistent, but without some way to create a state that is different than ambient, it's not a motor.

My understanding is that wax motors are based on actuation - they are intentionally and deliberately either heated to extend or cooled to retract, and if ambient temperature is used as one endpoint, the other will not be ambient (i.e. a heating element or cooling peltier).

Thermostatic valves are ambient at both sides - they expand and contract naturally with the existing temperature, so they are not motors per se, because there's no actuation beyond the temperature.

I don't know if that's necessarily true in practice the way it is in theory. If you knock down 95% of the issues they create over a certain period of time, in practice you'll have better code than a human would write, and it's all upside from there.

I've been using languages with stronger type systems and that's also a huge boon.

I think there's a pretty clear and wide gulf between "nobody could have seen that coming", or "I just wasn't on my A game", and frank negligence.

I once worked at a very large company, and one of my coworkers was like "eh, I'm not going to do any of this work, because I don't have to", and that was a clear indicator that the problem wasn't (just) with the system, but with him individually.

You stop relying on the agents following instructions exactly.

You need two pieces:

a) prompts, that tell the agents what to do and how to do it (and ideally, the why, where, etc, the full picture) - that's the positive half, that drives behavior the way you want it.

b) deterministic tooling that prevents negative outcomes, like linters, compilers, static analysis, fuzzing, testing, the more the better. This side should either be firewalled off from the AI or very carefully watched so that it doesn't drift.

The part that you put in the deterministic side is the "never do x" stuff - I have lint for long comments (which AI hits every single time it commits), all my dev scripts are in typescript, precommit hooks, massive CI, and I lint even for things like redirecting error to standard out, tiny stuff, and also e.g. static migration analysis so the AI never ships an exclusive full table lock in a migration, for example.

Typically, though, the diesel bug only happens in the interface between the water and the fuel itself, so if you have a proper filter and separator on the pumping system it won't be a factor.

Gasoline, when it goes bad, goes bad from top to bottom and is not recoverable through e.g. a centrifuge like https://www.youtube.com/watch?v=x4Fdu3e4gYw

Trust me, given how many people just accumulate trash because it's expensive, that's an absolutely sane and excellent policy. As a society, we want people to use trash disposal, because the alternative is expensive, painful, and disgusting.

This is Jevon's paradox - as you make the supply better and more efficient to use, people will demand more of it.

This same concept is also true of McDonalds historically - they own the land, lease it to franchisees, and realize the appreciation themselves. McDonalds are fairly economically valuable, so the capital gains by the time they're cleared and resold as development often dwarfs the value of the franchise fees.

My read, but the effective policies are infeasible, and the feasible policies are largely business as usual. Pretty much precisely as expected, honestly. I struggle to see how anything like this is going to be implemented without business as usual changing in ways that the current system is extremely resistant to.

Or encounter their CLI and it's lovecraftian combination of positional arguments, named arguments, piping, json documents, etc, distributed largely by random lot, as far as I can tell.