I believe the community will adapt and learn to leverage AI in their work. Same as they did with slide rulers, computers, calculators, etc. It will be a definite disruption, but I feel will ultimately allow people to focus on the more novel approaches, launch an AI task to follow all the implications and give a sanity check.

Large scale time zones are relatively recent and were only designed the way they are to accommodate easy math. Which now adays isn't so easy anymore given declining math skills. Even odds if I ask someone in NY what the time is in San Fran they'll pull up a world clock or do the math in their head.

Being synchronized to the hour is vastly overrated other than folks growing up that way claiming that's the way its always been. Frankly China had a good idea. One legal timezone to keep everyone in sync and each locality live under their own time. I'd hope the US would choose UTC but I'm sure they'd want to do EST if there was a nationwide one.

If it's permanent it's just the definition of the new timezone. Congress is too lazy to correct the time zones, it's just easier to change how DST is applied.

I had a wall world clock at the office. Showed like 6 different timezones around the world. SOO annoying they didn't allow firmware upgrades for it. Especially once DST got out of synch across the globe.

Parkinson's Law of Triviality published in 1957. Or in this venue more likely Poul-Henning Kamp's 1999 FreeBSD mailing-list post resurrecting it.

Probably has more to do with distance from the equator. Once you get to the tropics the days stay long enough that DST is not needed. Frankly less than 36 degrees latitude its not really needed.

Sounds like you need more sleep. If you still needed an alarm clock to wake you up an hour after your regular wake up time you are definitely sleep deprived.

Quantum mechanics is very well understood. Every particle in a configuration has an associated wave function. When squared that wave function now represents a probability distribution of where the particle will be when you collapse the wave function. You can set up configurations that create particles with a wave function with two distinct equal peaks. Collapse it like in a Stern-Gerlach experiment, and you get a completely random bit stream.

Nuclear decay is also completely random and driven by quantum effects. The time between decays in a sample is completely random and we know what the probability distribution curve is for most isotopes. That allows us to normalize it into the 50:50 random bit stream with some math.

Now getting enough throughput of random bits is an engineering issue and for virtually every case the $/ random bit per second has far exceeded cheaper methods that get you 99% of the way to true randomness.

Yes. Especially once SIP is turned off. But that's essentially reducing the security. About the only time I can see it being useful is if you refuse to get your kernel extension signed. So definitely worth it if you are developing a kernel extension. Even with it on you can run apps with self signed certificates.

The M series chips use a Unified Memory Architecture that effectively lets you use main system memory as GPU memory. The CPUs, GPUs, and Neural Engines all have direct access to the same memory. No need to be pumping bits between memory subsystems over PCI like the Intel ecosystem is still doing. Since most of these models are restricted to running on a certain amount of GPU memory you are either buying a high end NVIDIA board that alone is more than the price of a Mac mini or using the M series chips with extra system memory.

You want Replicators! This is how we get Replicators.

For complex issues I've found it to start circling the drain.

First answer, obviously wrong. I reply that doesn't seem correct, are you sure. LLM replies with kiss ups and says I'm so smart for checking and it's obviously wrong this is the correct answer.

Second answer looks self consistent and seems to solve the case. But after going moving forward with it see that it's an algorithm extremely fragile to edge cases that causes it to blow up. Point out the cases and instead of fixing the core issues it comes up with solutions for each unique edge case I present where it fails giving a massive amount of spaghetti code.

Third push and I get the first answer back again.

The biggest value I see is that I'm very well informed about the issue now and code the thing myself in a manner that is efficient and resilient to edge cases.

Honestly it reminds me of grad school when I had to grade undergraduate papers. I spent the bulk of my time trying to figure out what rationale the student used to come up with the wrong answers in order to figure out how much partial credit to award. It gave me a very thorough understanding of the core concepts. Still remember one hard mid-term where every student's answer was different to a problem that only had one solution. Had to dig deep into all the concepts of that question only to realize that the subtle calculations of production efficiencies were a moot point since the parameters specified in the question would have caused a massive explosion in real life obliterating all the products.

The fine print is its an impurity. For RTG's you want the short half life of Pu238 but it's not worth the lift to do isotopic separation. They just use production methods that primarily produce Pu-238 but there are very low probability paths in those production cycles that still make some 239.

Pu-239 is everywhere due to all the detonations that have been done throughout the world. The average person urinates a few hundred million atoms of Pu-239 every trip to the bathroom. So yes Virginia, your body contains "Plutonium isotopes including Pu-239".

A closer argument would be for the Sony Playstation and MS Xbox. They both charge 30% of revenue for app developers. The exact same arguments can be used against them.