I think you're absolutely right. What makes me so disappointed is that Gizmodo threw out accusations of AI hallucinating without bothering to follow the article's Wayback links that provide positive proof. That's yellow reporting at it's worst, and is entirely misleading.

Looking at my Microsoft Authenticator app right now, I see dozens of accounts based on the same email address, and not a single account that has an empty provider. I have page after page of accounts in here, and would love for better organization, but I'm yet to encounter a QR enrollment scheme that has an empty provider field, and if the article is to be believed, I would have to run into that TWICE before there would be a collision. I kind of agree with Microsoft here. If the QR doesn't have a provider, that credential is garbage to begin with, so overwriting it should not be a big deal. How would I even have any clue which credential to use without a provider name listed with it?

Not sure if you're aware, but Columbia University already DID move off it's expensive real estate and onto a less expensive campus. The land that Rockefeller Center is built on is still owned by Columbia, and their lease on it goes a long way towards paying their bills. This old report shows Columbia owns 246 sites within NYC: https://ny.curbed.com/2018/9/1... NYU is another huge land owner within NYC, and I suspect that more universities than not have their investments tied up quite heavily in real estate. Your argument that the land underneath these universities is simply too expensive to waste on learning ignores the fact that the property values came about largely from the universities creating that value. Can you even imagine an "XKCD: What If" on Ithaca if Cornell were erased from the map?

My hovercraft is full of eels.

I don't think PSI means what you think it means. CPAP pressures are measured in cm-H2O. 20 cm-H2O = 0.284467 PSI

> it is hard because it is so small, H2 is just 2 proton and electron cloud pair, it's so small compared with everything else

NO. Helium is smaller.

Just eat some french fries, chips or peanuts. Your fingers will be sufficiently salted.

It is either based on MOSFET or more likely IGBT, but requires a lot more in the package to actually work. The problem is that while they've claimed to be able to be able to have it act like a circuit breaker, it will not. tl;dr: there's no innovation to be found here, just smoke and mirrors.

Sure, you can turn power on and off with a solid-state device. But for starters (and this was glossed over in the article), there are significant losses due to this, where a mechanical switch has almost no loss. Using such a device at every step in your power path would be grossly wasteful of energy.

Next, the claims that this switch can act faster than a mechanical breaker are disingenuous. The time limitations on a mechanical breaker are largely based on the fact that it is interrupting an ALTERNATING CURRENT circuit. The voltage crosses zero 60 (or 50) times each and every second. So, at best, on a magnetic trip, a mechanical breaker requires one cycle to trip (roughly 17 milliseconds). But thermal trips can (and SHOULD sometimes) take hours, because circuit breakers are protecting downstream wiring, which has a real-world duty cycle. i.e. UL already lists existing mechanical circuit breaker technologies, because where they are used, faster fault clearance is not required. In the situations where faster clearance and higher fault current interruption is needed, we still see fuses, and I highly doubt that this solid state device could replace a fuse.

Why not replace a fuse? Because interrupting an poor power factor load can lead to the near instantaneous liberation of enormous amounts of energy. The energy has to go somewhere, and the voltages created when interrupting a fault on an inductive load will spike to the point that any solid state device will suffer damage. So, now your IGBT will need zener diodes and MOVs protecting it (and those MOVs NEED fuses, and this all adds to the energy losses). And yes, you could try all this, but I still doubt you could beat the performance of a fuse, even in a package 20x the volume (if it is at all possible).

But wait, aren't there already circuit breakers with microcontrollers? YES! They still use mechanical contacts though, which slows down their action, but that's acceptable, because electronically controlled circuit breakers are actually chosen when you want to SLOW DOWN the breaker speed. Wait, what? Yep, slow it down. Breaker coordination is a tricky science, because the full current of a downstream fault will be seen at an upstream main breaker too, but suffice it to say that you do not want to trip the main when another breaker should clear the fault. So you may choose upstream breakers that hold off their "instantaneous" trip for five cycles or so (83ms).

So, what's the application for this? Well, my guess is they're trying to make a "static switch" without a mechanical contactor in parallel (which is just plain dumb IMNSHO), with the idea that it can be used to switch power sources from wind to solar to grid as an example, without interruption. But solar inverter grid-ties already do this (using the same solid state electronics), so again, I don't see any innovation in this article.

A real world example from an orthodox rabbi I know was that he drove to the emergency room on Shabbat, but with his obligation to drive over at the point that he arrived, he left the car open and running right there at the hospital entrance.

That's why I always wear my full body Sasquatch suit underneath my clothing. I just have to be careful when going out naked in the woods to stay away from hunters and always make sure any photographers don't focus on me.

Clouds would seem to break this solar navigation system just like any other.

Also the eventual fate of any cloud service that is ostensibly free, or paid for up front, relying on future sales, without any recurring subscription charges to support its long term maintenance.

You do realize that the moon moves across the star field. It was probably just a distant star that you saw "move" in relation to the moon.

Wensleydale, is that you?

There are roughly 365 days in a year. What makes you think that that SINGLE day is representative of the rest? In areas where air conditioning is a big thing during the summer months, the peak will be in the early afternoon on warm days. What I'd like to see is a time weighted annualized curve of this data, per region.