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.

Overwhelmingly? The only link that had real meat to it was the money laundering one and event that had the big detractors putting it at ~ 1% of transactions where as the formal surveys still put it at 0.25%. Fiat money is estimated to be used at 2-5% of all transactions.

The bill's draft text is at https://www.legislature.mi.gov....

It defines: "Circumvention tools" means any software, hardware, or service designed to bypass internet filtering mechanisms or content restrictions including virtual private networks, proxy servers, and encrypted tunneling methods to evade content restrictions.

It also has this text:
An internet service provider providing internet service in this state shall actively monitor and block known circumvention tools.

So VPNs must be banned by ISPs if the bill passes.

Even if they start doing deep packet inspections to block the protocol just wrap it in SSL on port 443.

It's a learned skill and can be precise to the minute over at least a day. Granted it's gone by the wayside since there are so many exact reference's available now. There's no need.

I'm confident I could do 30 min margin of error over a few hours and I know many other much better than me. If I have an appointment in the morning I'll consistently wake up exactly 1 minute before my alarm no matter what time I go to bed.

Context matters. That was a wrap up/summary message in her broadcast. She was reporting on early data from the CDC that showed 90% efficacy against the early strains. In the vaccine world with the basic reproduction numbers associated with early COVID anything over 70% efficacy would stop the spread. Reporting statistical results to the scientifically illiterate public is always a challenge.

By far the worst messaging was with the face coverings and 2m spacing. Unfortunately they were based on studies of airborne Typhoid transmission. When they finally did the science for COVID-19's airborne particulates it was closer to 10 meters was needed to get to the 95% confidence level. Typhoid is transmitted with a bacteria of much larger size so fell out of the air quicker. Granted it did help because airborne viral concentration rapidly decreased with distance but it was messaged that you'd be safe at 2 meters.

Biggest reason is less regulations. Currently they need to get money transmitter licenses from every state and territories which all have slightly different requirements. Wise can't even legally get one in every state because some states require US Ownership. By getting a federal charter they can leverage that across the entire country. Don't need to worry about compliance to 50+ jurisdictions.

There are also the analog chips on the horizon. They don't have the precision in calculating the weights of a ML matrices but they are good enough for the algorithm needs and use a fraction of the power.