lambdaone

Deep, deep down, the average person wants to be controlled and told what to do, just so long as they don't have to acknowledge it to themselves. Clinging to the garment of a Daddy or a Mommy or a god or priest or a Great Leader is the usual way to do this.

Buying brands that advertising has told us will make the anxiety go away, or equivalently believing ideologies that propaganda has told us the same, is another.

Note that I count myself among this number - I'm not holding myself out as a superior free-thinker, I think it's likely that I'm just as unaware of my personal flavour of self-deception as anyone else.

Clinging to chatbots is just a new version of the same thing.

"Mental gag reflex" is exactly right. I'm running two different instances of Claude Opus 4.8 on xhigh right now, and I'm absolutely fine with it because that's what _I_ want to do.

AI features on my toothbrush, toaster, refrigerator, doorbell, washing machine, word processing software, TV, whatever, without my actually asking for them first are THINGS I DO NOT WANT, and adding those features to those devices will cause me either to have go to great lengths not to use them, or - much more likely - just not to buy them at all if I can.

Exactly this. I want a system that can understand basic instructions like this, in context, with all the necessary unspoken assumptions implicit in natural language and the problem domain, and get it right _every single time without fail_. A bit like AppleScript, but in natural language and smarter.

Most definitely _not_ a cleverer system that sometimes hallucinates or goes off on tangents, or sends my personal information to parties unknown.

I've written my own autoplacer/autorouter. Placement is where you put the components on the board, routing is how you shape the traces to interconnect them.

It does a pretty decent job on small hobby-project boards of ~40 components (which is my use case at the moment), and I'm working part-time in the background on scaling it further.

The resulting designs pass all the KiCad electrical and geometry checks. Granted, I've spent about a year working on this problem, and it's hard, but not that hard a problem, providing you can avoid falling off the exponential cliff by decomposing it into hierarchical subproblems.

Quick-and-dirty unsupervised whole-board synthesis from schematic takes about 5 minutes, longer if you want cleaner output with nicer-looking better-routed traces.

As others here have said, placing is the real problem to solve, and that's where the magic happens. Place the components right, and routing is a relatively easy loosely-coupled constraint programming problem, place them wrongly, and you will have to get used to seeing the word UNFEASIBLE in your log output.

The answer to grandfathering things in is not to give them an indefinite examption from the rules; instead, give them, say, a 10 year period of exemption, giving the owner enough time to fix the defect and to spread the expense of doing so over time. It's not perfect, but eventually everything ends up being fixed.

"Probably the wrong framework for most use-cases" is not that much of a recommendation.

SPAs are a monster that grew and grew; an idea that at first seemed like a clever optimization has generated huge and complex systems that are a nightmare to work with, and are bloated, slow and insecure.

Web components are the next big idea. I hope they have a chance to work.

There is a medical treatise, "On Assholes", just waiting to be written here, in much the same way that Harry G. Frankfurt introduced us to the technical academic concept of 'bullshit' in his book "On Bullshit".

There's not much information about transfluxors available, although a patent for them goes into quite a lot of technical detail, and is referenced in the Wikipedia article.[0] If anyone knows more about this, the Wikipedia article is definitely in need of expansion:

[0] https://en.wikipedia.org/wiki/Transfluxor

This is truly disturbing, and trying to sneak it in like this without public discussion is disingenous. Hopefully it will be shot down like last time - at the very least, there are surely antitrust issues here.

There are so many ways around this - for a start, you can use some other working fluid that boils at a lower temperature. Or you can choose a different thermodynamic cycle that doesn't involve phase change.

I once aspired to American citizenship, and was dazzled by its wealth, opportunity, can-do attitude and freedom. Now I can't imagine wanting to go there - everything I see or hear, from both American and other sources, right or left, suggests a deeply unhappy country at war with itself.

It seems obvious to me that there should now be a concerted and open effort to detect malware in supply chains based on AI-based scanning. Sure, there will be an arms race in malware obfuscation, but that was coming anyway. Manual review is useless at this scale - it is just not happening.

At today's prices, that's around 1400 tons of gold. It's made out of lots of individual gold bars, and they can certainly move it in increments, by road then air or sea, unless the American government stops them doing it.

People ship million-dollar assets all the time. It would be a huge task to make 160,000 such trips, though.

Very insightful. One key sentence sums it up: "He shipped a product, but he didn't learn a trade."

This is going to get worse, and eventually cause disastrous damage unless we do something about it, as we risk losing human institutional memory across just about every domain, and end up as child-like supplicants to the machines.

But as the article says, this is a people problem, not a machine problem.

False positives can be eliminated mechanistically by testing if they actually work, in a sufficiently isolated automated test apparatus.

The hard thing is reducing detected crashes to well-formulated test cases that help rather than hinder maintainers.

What a fantastic company HP used to be, back in the day. They led the way in scientific equipment and calculators, and even desktop computers for a brief moment.

They even made PostScript laser printers that were built like tanks and were a by-word for reliability.

Now they are just famous for being the printer brand everyone hates, and this is just scraping the bottom out of an already empty barrel.

The article says "I don't know where this fascination with getting everyone to download your app comes from."

The answer is really simple and follows on from this article; the purpose of the app is even more privacy violation and tracking.

This company sounds like it has months to live, or until the VC money runs out at most. If this idea is good, Anthropic et. al. will roll it into their own product, eliminating any purpose for it to exist as an independent product. And if it isn't any good, the company won't get traction.

Of course, the trained model they use to do the code generation may itself have been trained on the very open source code they are trying to replicate 'cleanly'.

Amazing hyperbole, and a deflection from the real issues. You can fight against wrongdoing without actually advocating people being killed.

Right now, climate change is an undeniable fact, its causes well-known, and the evidence for it now part of everyday life. If anything, its effects have been underestimated to date, and 'non-believers' in it are either fools or acting based on morally repugnant principles.

Yes. The obsession with demonizing AI/data centre loads seems to be a deliberate distraction from the much, much larger carbon loads of the economy at large relative to which IT power consumption is a tiny proportion.

This is terrifying, and those fighting against stopping or reducing global warming should at this point be regarded as hostis humani generis

This is a great description of something I've been thinking about in terms of concepts like regression to the mean, clustering and median filtering - the space of LLM output is much smaller than that of the input, precisely because the LLM works so hard to extract minimal-information patterns from its input.

It's a sign of how far we've come that this article says "Break Free from Google and Apple", not "Break Free from Google, Apple and Microsoft".

None of this required a police state. Just people working together to cross-correlate information in the way that you would expect to be able to do in an open society.

What wrong do you think was done here? What would you prefer to be different?

A large industrial robot running an insecure protocol - what could possibly go wrong?

You can use ssh with the None cipher, thus disabling encryption entirely while still using the rest of the protocol.

It's certainly one way to do arena-based garbage collection.

Multistory DCs are commonplace in major cities.