Companies should, in general, avoid keeping any user data one second longer than they legitimately need it. Data should be viewed as a liability more than an asset, and every company that achieves significant scale with user data soon realizes this, though some take longer than others to learn the lesson.

The reason it's a liability is because it exposes the company to an endless stream of warrants and subpoenas, and contrary to popular belief, the government does not pay for that data. In some cases the government might pay a portion of the expenses the company incurs in responding, but it's always still a net loss, and most of the time the government just shows up with a court order and says "Hand it over or be held in contempt of court". What makes it even worse is if companies actually want to be good stewards of the data and not hand over more than they should, they'll often get overly-broad requests and then have to go to court to fight to narrow or quash the requests.

This all gets very expensive. It's just a cost of doing business... but it's a cost that can be avoided if you don't have and can't get the data.

Encrypt everything, and arrange to destroy the keys when you no longer need the data is the best way to do it, too. Scrubbing large databases quickly and effectively is hard. Deleting a set of AES keys is a lot easier. This also means that you can focus your internal access controls on controlling which systems and employees can get the keys.

I'd encourage Ring to find ways to close the window of access even more.

~nt~

The quote was from the Bavarian government guy, so it seems unlikely that he's concerned about Waymo's competitive worries.

Also, German regulators are very cautious, so I expect Tesla's camera-only and less-mature robotaxi service will have a harder time getting approved.

Well, given that Waymo has been testing for nearly 20 years and been operating commercially for six years, currently operating nearly 4000 vehicles in 11 cities doing 500k rides per week... there's not a lot of risk.

Roads in Munich aren't terribly different from roads in US cities. Stoplights work a little differently. Munich does get a little snow, but rarely very much and it almost never lasts long. If Waymo's vehicles can't deal with it they may just suspend service... though I expect they'll probably handle it fine. They're supposed to go live in Detroit and Minneapolis this year, and if they can handle the weather there (where just shutting down for snow is not an option), Munich will be no problem at all.

So, not much like OceanGate.

The recording engineer only records (and maybe mixes) the audio that is generated, which is part of what I meant by "mechanical process", along with generating the audio. Dre could use AI to transform his ideas into digital audio tracks, obviating the recording engineer and the performers, and to then select and mold the results. This would still require talent and creativity, which was my point.

Caldera didn't buy SCO.

I didn't bother getting into the Novell situation or the fact that TSG didn't actually acquire the assets they'd have needed to do what they tried to do, since my goal was just to distinguish between SCO and TSG, but, yes, the details get pretty intricate.

It's almost certainly all operational expense, same as the salary paid, which means it's "written off". But "write off" doesn't mean "it's like you didn't have to pay it", it just means that it's subtracted from revenues when calculating profits. Net is that writing something off saves about 21% (the corporate tax rate) of the amount written off.

It's the same as your personal taxes, more or less. When you donate to charity, you get to subtract the donation from your adjusted gross income. If you make $100k and donate $10k, then it's as if you made $90k. Supposing (for simplicity) you pay a flat 20% tax rate, and you didn't make the donation, you'd pay $100k * 0.2 = $20k in taxes. Having made the donation, though, you only pay $90k * 0.2 = $18k in taxes. You donated $10k and basically got $2k of it back from Uncle Sam so the donation only cost you $8k out of pocket. But you still ended up $8k poorer (but the charity is $10k richer!).

The principle is the same with corporate "write-offs", though the mechanics are different.

Silicon Valley companies will pay the $100k. The positions typically pay $500k+ (base + bonus + equity + benefits) and companies are accustomed to spending tens of thousands of dollars in addition to deal with the H1B filings, and for other immigration services, including fees for green card filings, etc. An additional $100k is painful, but since it's good for 6 years, they're already paying around $3M, so it's only a ~3.5% increase.

I used it quite a bit and found it pretty decent. I never used a modem, though, so you may be right about that.

Except time and willpower. However, the McDonald's analogy does hold up -- nobody is ever going to get rich flipping burgers at McDonald's, precisely because it requires so little skill to do that job, making McDonald's workers completely fungible. Since anyone can man a grill, no one is going to command a salary much more than what a high schooler will agree to do it for.

Many require you to use the phone app for MFA, even if you're using a web browser.

I guess there's the option of just doing all your banking in person, at a physical branch location... or at least an ATM.

Assuming for the sake of argument that what Dre does is creative and that he is talented, very little of making great music is about the mechanical process of executing on the generation and recording of sound. That's important, but there are a lot more people who can do that than create something that is new, and good.

Creation (in any medium) is about the ideas, and about the ability to recognize what works well and what doesn't. Given that, I could completely see AI being a useful tool for a creative, talented artist. They have to supply the ideas, in the form of prompts, and they have to recognize whether the results are good. The AI is just a time-saving device for the middle step of generating representations of the ideas.

The economic problem with having an AI produce music for you is, you never get the opportunity to learn how to produce it for yourself. At the end of the day, you're paying them $200/month for the privilege of remaining unskilled, and you're competing against everyone else who knows how to enter "write a catchy song for me" into a prompt box. You can't make money that way, because the moment you do, the AI companies (who own all of the talent and all of the production machinery) can simply appropriate your income stream for themselves.

It's the difference playing Guitar Hero and taking guitar lessons -- both let you feel like you're a competent musician, but only the latter will ever get you a career as a musician.

We reached peak woke four years ago. DEI is dead. The corpse is still twitching in a few corners of academia, but that's about it.

I don't think the center left ever moved. The center left just didn't push back against the far left because why alienate them? I mean, the far left is perfectly capable of just deciding to sit out the elections. A non-trivial number did exactly that in 2024 because they were mad that Biden didn't take a hard line with Israel. And given that the GOP has turned insane...

IMO, the center left does need to push back against the far left. But I don't think the mainstream Democrats are going to get that message until they get smacked in the face with just how out of step a lot of those far left positions are. It's not going to be DEI/woke crap now, it's going to be the full on socialists.

But the only way for the voters to deliver that much-needed smack is by handing them a severe electoral loss. Which also isn't going to happen because the GOP has gone insane.

Likewise, the GOP needs to be smacked in the face with a solid loss to bring them back to reality. But that's not going to happen because DSA...

I think for the next cycle or three, we're going to be stuck with two increasingly-extremist, increasingly out-of-touch parties, and neither is going to get the kind of big loss needed to wake them up.

People don't do that. Even if they come to their senses the vast majority will just not want to talk about it... or else they're revise history and say they never actually supported what they support.