Given that the Roberts Court is one of the most corporate-friendly in history, this decision comes as something of a surprise.

Nonetheless, it appears to be largely concordant with the so-called "Betamax case" from the early 1980's which established the principle of significant non-infringing uses as a defense and, despite passage of the DMCA, still largely informs the contours of contributory infringement.

So you think there's a commercial market for missiles that fail in flight 90% of the time? You believe that they would engineer missiles with a GPS that couldn't handle the speed? You don't think that China, who boasts their own GPS-like network of satellites and builds their own receivers, can't build a receiver that works at Mach 5? You don't think that they're capable of building a dead-reckoning system that can land within 50m of target in the face of GPS jamming? You don't think that the country that's likely to land on the moon in the next 5-10 years can build a rocket body that can manage to stay in one piece?

If they're gonna sell them, they're gonna have test results showing that they work as expected in a benign environment. Whether it's 99.9% success, 99%, or 90%, there'll be real numbers based on real test launches. The people that they're selling to ain't gonna buy a 10% success rate missile, but they might buy a 90% success rate missile if the price is low enough, and everybody would be happy with a 99% success rate missile. And remember that their initial customer is likely to be mainland China, who has ways of dealing with disappointing suppliers.

Well, any refractory ceramic would probably work fine. The difference between an "aerospace grade" refractory ceramic and a "commercial grade" refractory ceramic is probably a 1% difference in missile failure rate at a 100x increase in cost.

Let's do the math:
Using aerospace grade parts, you launch 100 missles, 90% of them get intercepted before reaching your target, so you get 10 strikes.
Using commercial grade parts, you launch 100 missles, 10 of them fail during launch/flight, 90% of the remainder get intercepted, so you get 9 strikes. But, because the missles are 10% of the cost of the aerospace parts, you're able to launch 10x as many, so you get 90 strikes.

Where this leads is terrifying.

>>> this will result in software with obscure bugs that are near-impossible for a human to find or fix because no human even understands how the software works.

And that's different from the current state of Software Engineering how, exactly?

ehh, I've had the opposite experience. I've got way too many years writing C for embedded systems, but needed an Android app. So I asked ChatGPT to create an Android app for me that would do MDNS and Bluetooth discovery, pop up a dialog to let me choose from the discovered devices, and then connect over WiFi or Bluetooth as appropriate. And, after spending an hour downloading and installing the Android toolset, the program compiled first time and did what I asked. I did my normal step-through-line-by-line verification, and it was fine for prototype code.

As someone who'd never written an Android app in his life, this was eye-opening. It would have taken me weeks to get there pre-Google, and several days to get there with Google, but about 10 minutes with an LLM. Now, would I trust it to write production embedded code? Ya know, with the appropriate LLM-generated tests and human validation, I think so.

My son recently graduated with a degree in Software Engineering; I try to tell him that the future for that is going to be what we today would call a software architect or designer. The Software Engineer will be judged on how well they can write the specs and requirements fed into an AI to generate code, not how well they remember the details of how the C++ Lambda function works, or be able to generate a complex regular expression -- and be able to read it six months later. We're not there yet, but there's so much money being poured into the problem, and it's such an easy-to-see evolution, that in 10 years the concept of actually paying attention to spaces and semicolons will be quaint.

I'm assuming that, on a cosmic time scale, the fragments will re-coalesce under their own gravity?

OpenAI is amending its Pentagon contract after CEO Sam Altman acknowledged it appeared "opportunistic and sloppy." [ ... ]

Well, if there's anyone who would know about slop...

Sounds like Micros~1 doesn't want to deal with actual people, much less the consequences of their own boneheaded decisions.

Of course, if Discord had a backbone (and ethics), they would summarily remove the filters, and smack Micros~1 for making them look bad. And if Micros~1 gave them any back-talk about it, they could reply, "Well, it sounds like you should set up your own rules on your own globally accessible chat network. I hear you already have something along those lines. Something called... Teams, I think?. Knock yourselves out..."

The argument proffered by management appears to boil down to nothing more than, "Well, everyone else is jumping off the Empire State Building, so what's your problem?

Also: These lemmings are in for a FAFO-fueled rude awakening when they discover all the slop they've checked in and shipped/deployed, being machine-generated, is uncopyrghtable. "Um, actually... It's just like using a C compiler, transforming the programmer's intent to runnable code, so..." *SMACK!* Wrong. Compilers are deterministic. You can draw a straight line between the source code (and therefore the programmer's creative choices and intent) and the resulting binary and, given the same input, will generate the same output every time (indeed, if you do get different output, it's a bug) LLMs are anything but -- they'll give you different answers depending on what you may or may not have asked before, the phase of the moon, and which vendor paid to have the LLM preferentially yield responses using their commercial framework.

In short, this is a bone-headed move, and when it came time for the managers' performance review, I'd give a negative score to anyone imposing mandatory LLM use.

So a bunch of companies took a bunch of educational time to teach a bunch of children skills that were obsolete before the children even graduated from high school? And now they're pivoting to teaching this weeks new tech fad with no concept for whether or not it'll be obsolete in a decade also? Are there consequences for this kind of antisocial behavior?

Reading, writing, thinking, math, finance are basic skills that every child needs to have IMHO. They aren't made obsolete by a where the tech bros are investing their billions today.

It's been fairly well established over the years that anything electromechanical (phono cartridge, speaker, headphone) is by far the biggest variable in a high end setup.

Moderate and high-end preamps and amplifiers are all pretty similar at similar power levels. You're trying for a flat amplitude and low phase shift over a 3 decade low-frequency range - that's not rocket science. Moderate cost pre-amps/amps will tend to have a higher noise floor (more background hiss), but otherwise they'll be remarkably similar.

But the speakers/headphones - ah, those are magic. I went to a private engineering college with a number of people who had more money than sense, and got to demo speaker systems that varied in cost by a factor of 100. In general, the most expensive ones sounded much better than the cheapest, and a little better than the middle ones, but even so there were noticeable differences in reproduction amongst the highest end ones that didn't exist amongst the purely electronic components. I couldn't tell you which of the high-end systems were best, but I could tell you which of them I preferred.

Or until speaking out against the government gets you labelled a "domestic terrorist", and the Department of Homeland Security decides to crack down on Domestic Terrorists, because who isn't in favor of that?