They're also known for sponsoring mathematicians and their channels, notably Numberphile and Matt Parker's StandUpMaths channel.

They are recruiting mathematicians, now it makes complete sense why they have summer camps and such.

With all the known compromised machines to analyze, why can't Apple figure out how this spyware got in and plug the leak(s)?

There aren't many. These attacks generally only affect a few hundred people. And Apple is warning people because they see strange access patterns coming from their phones to their servers.

Apple doesn't have the malware or the devices in hand to analyze them (and fix the holes) - they only noticed that some odd access patterns were happening and notified users.

To track them down Apple would need a copy of the malware and likely the events triggering it (e.g., text message). They likely have none of that. I mean they probably could, but that would me having people to surrender their phone to Apple for analysis, which also means giving Apple all the data contained within. Chances are good that the people involved would be wary of what happens to their data.

So no, Apple can't go after them because Apple basically doesn't have the malware to analyze. And the people who can give it to them are likely people who can't give it to them.

Are you sure you want to open the door to suing people contributing to open projects?

Generally there's a difference between contribution and vandalism.

Now, maybe it was harmless - or maybe it leads to real damages (e.g., someone packs up their family and heads off to the fictional beach)

Given how unhinged some people are, it could even result in fatalities if someone went there expecting a beach then finds people telling them there's no beach and it's just a video game.

I suppose "almost" is the key word there. Ask astronauts if "good enough" is acceptable for the software for the rockets, space stations, etc. Airplane controls and car control (in the future when we get to level 5) are other places. There are times we do want perfect, and "life dependent" software definitely qualifies for that.

Not really. The space shuttle software couldn't handle year changes, for example, which meant no space missions could be done near the end of the year.

Is that software "good"? I'd argue not - because date handling is basic enough, and years are often enough that software generally has to handle it as a common enough event.

But it IS "good enough" - how many missions are going to be done near the end of the year? And would the effort used in making the software handle the year transition be better spent doing something else?

What about level 5 autonomous driving? Does the object avoidance system need to know everything about obstacles? Or would it be adequate if something wasn't in its database and it was just treated as "object to avoid"? Just seeing things on the road as "things to avoid" versus "that thing is a bicycle" or "that thing is a pedestrian" is also an example of "good enough" - you have to avoid both, and knowing the difference from one another might not be terribly useful - both are objects moving with a certain speed and you can project out if their path in the future may intersect yours. But knowing that the object is a car, a bicyclist, or a pedestrian isn't as useful as say, knowing its speed and the chance for it to randomly change directions (generally speaking, faster objects can't change direction very quickly, so a slow moving pedestrian can suddenly walk in your path even if they were going parallel to you without warning, while a cyclist will require some turning radius).

The problem are the tokens are generally not as portable. I'm still trying to find one that lets me install it in multiple places.

My bank requires it, which is inconvenient because I'd like to do a transfer, then realize my phone isn't near me and have to run it to in order to sign in using its 2FA system.

The company I work for did it for Office365, which means if I need to log into Teams on the web, I have to run to phone to authorize the login as well.

But since I can't authorize another device, i'm stuck with running back and forth to the one device that has the keys. Which may mean I might be out and about then decide I needed to do something, but realize I'm nowhere near my phone to do it.

I don't mind 2FA, but please realize that "something you have" may have multiple things of which they may not all be convenient to access at all times. So how about letting us register multiple devices?

This was made even more complex when I switched phones and it wouldn't let me deregister the device, but wouldn't let me register a new one either. I had to disable 2FA, deregister the old device, then enable 2FA which then let me register a device.

For an *audiobook*? It's pretty cheap, actually. You can buy a good microphone with a built-in DAC for about $100 ($200 tops), plug it straight into your laptop, and use any number of free programs for editing (and for whatever minimal post-processing you might want to do). The only other requirement is a quiet place to record.

You haven't browsed the videos made by new YouTubers, have you? They do that and it still sounds terrible. It's not equipment, but skill and effort in obtaining a good mix. Skills that are quite different from writing a book.

Audio problems are prevalent on YouTube, and you'd think people would know better. And these are short videos. An audio book can run 12+ hours, maintaining a proper levelled mix is a lot harder.

Is there a point to using stablecoins? To me, they seem like Showbiz Pizza tokens, where one trades a currency that is known and good (for better or worse) for someone else's in order to play in the arcade.

As a store of value, if they are tied to the dollar, why not just use a dollar, and be sure the rug doesn't get pulled. As a method of trade, they might be useful, but getting one's money out of the currency may be difficult to impossible.

I believe they're just another form of gift card. You spend useful currency on something that provides a similar value but in a less convenient format.

So these could be anything from store gift cards to others like Visa/MC gift cards.

They're just fancy gift cards.

The future cost of IPv4 addresses seems like an interesting economic study. If the price goes up a lot then that will encourage some companies to sell off part of their ranges and move towards more NAT and/or IPv6. Also if IPv4 addresses get costly then cloud providers will start changing more for IPv4 access to encourage IPv6 instead. On the other hand, people might pay more so they don't have to migrate a working setup.

It's like oil - the price of oil rises, then the whole supply chain starts coming online to reap the profits, while drivers start the migration to alternate fuel vehicles (AFVs), which includes things like EVs and hybrids.

But you don't want the price to go to high all at once lest you start losing your customers (who can switch, or decide to not drive as much). But you also don't want the price to drop because it's profits.

Gas prices are never coming down because of it - no one wants prices to drop, so new sources of oil are being held in reserve but not extracted.

It's why peak oil theory is incorrect - while gas demand is relatively inelastic, it's not perfectly inelastic - so the goal is to maximize profits while minimizing alternatives - to get people to not switch to EVs, and likewise to not change their routines to save gas. But at the same time, as prices rise, suddenly harder to extract forms of oil become profitable so to supplement supply.

Consider that we've sold the last IPv4 address a long time ago (over a decade) yet we're still no closer to switching to IPv6. IPv6 is being used (LTE and 5G cellular networks are IPv6 only) but everyone is still trodding around using IPv4. Heck, some cellular providers don't provide IPv6 access.

Talk is cheap. The reason you would never do this is the exact same reason nobody else is going to do this.

Sounds good, let's see it IRL. How much usable energy per unit of battery weight?

Don't know about weight, but you can buy 18650 cells using Na-Ion right now. They have the power capacity and curves of LiFePO4 cells at the moment.

The key part is that we have tons of sodium, unlike lithium, and a lot of it is already in ion form. Earth's lithium supplies are limited, while sodium supplies are basically limitless, and thus, it's stupidly cheap and unlikely to rise due to its abundance.

Sodium batteries are very similar to lithium, since it's in the same group (one row down) so the properties are similar. Hopefully that means enhanced sodium cells are soon as they apply the advancements made to lithium batteries to sodium batteries.

But you can apparently play with them today. A video on a YouTuber playing with them - https://youtu.be/s6zcI1GrkK4

Florida while always being close will likely be red for a few more cycles at least. Shifting demographics should be in favor of making things closer but the Florida Democratic party is very ineffective in my opinion.

Not likely.

Florida is home to basically all of the US billionaires - there are very favorable tax laws in place and basically the centi-millionaires and billionaires are snapping up properties in Florida.

So no, they're not aging out, rather, that's where the concentration of wealth is happening.

Of course, one should also note this because when the proles start to rise on the rich, they would be concentrated in one state, making it rather easy.

It's entropy, plain and simple. Sooner or later, no matter how secure an organization may be at any given point, skip ahead a few cycles, and attention to detail wanes. Managers stop asking questions, project leaders reprioritize thinking the problem is solved, staff do a "monkey see, monkey do", and then new gaps open up, get taken advantage of, management go into a state of denial, project leaders can't get their teams to give a damn, and then the inevitable breach or audit reveals the extent of the vulnerabilities, and management sends out the big press release that's always "We're reprioritizing security because we take security SERIOUSLY!"

Rinse, repeat, endlessly until the heat death of the universe shows entropy is always king.

Anytime Spotify loses customers, it's the fault of Apple and Apple Music stealing customers away from Spotify.

Spotify can only be the #1 streaming music service. Any loss is due to Apple. There is nothing Spotify can do that causes it to lose customers without it being Apple's fault. Don't implement a feature user wants? Apple's fault.

Are they going to ban license plate reading/lookup too if it's run through Azure? Traffic cameras? This is outside Microsoft's decision making IMHO. They either accept money for services rendered or they don't. Discriminating against US law enforcement should not be an option and should actually be illegal (again IMHO). We're supposed to trust the government. If that's an issue, then fix the root of the problem. Disallowing police from using technologies is not the right fix.

That's probably it. First, using generative AI tools might add details to police reports that were not present, and this has very problematic issues regarding evidence and testimony. We know AI loves to hallucinate, and even worse, do it without telling you. If you're writing up a description of a person you don't want it to suddenly embellish the description with details that weren't there.

Likewise, using it to generate an image of a person from a description might add details that aren't there, and even worse, may falsely incriminate someone because the generated image looks like someone familiar who is totally innocent.

Same goes for image recognition - we've already had people false arrested over facial recognition failures - of which if only someone actually compared the photos from surveillance cameras against the person themselves, would've told you it's the wrong person.

Microsoft is basically saying AI is too unreliable to use it for anything that could affect someone's life. Whether it's because the training data set lacks diversity in its images, or other things, it's not in a position where you can seriously affect someone's life with it.

Having software accuse you of committing a crime is something Microsoft wants to avoid because the technology is not there yet. Giving it an image and asking an AI to identify a person is not something AI is ready for yet.

And honestly, the technology just isn't there yet. The output of AI should only be used in a fictional context, You can't trust AI right now in ways that could seriously affect someone's life.

Well, the DMA basically says Apple can't prevent people from installing apps outside the App Store. That's about it. You can choose to use the App Store or not, but Apple cannot control the content or distribution of apps to their devices.

Which is why Apple did comply - all apps still need to be signed by Apple, but those in the DMA area can use an alternate signing mechanism where you submit your binary to Apple, and Apple signs it. They do basic automated checks but otherwise do not review it for content. So if you wanted a porn app, you can - Apple will sign in.

Likely that app signing will include an identifier because the app will be linked back to the developer - so if someone distributes malware it can be traced back to the developer that created it.

The DMA didn't say anything about developers remaining anonymous. All it said was developers have a right to have users install their apps without Apple preventing it. Likely the whole thing of Apple holding developers accountable keeps the platform clean and honest.

Of course, it does mean developers have had signing keys stolen because malware writers have had to exploit developers to get malicious apps signed. I mean, Realtek got their signing keys stolen and used by malware. There's also reports of an Apple developer getting their signing certificate deactivated because someone stole it from their Mac and released a bunch of malware with it.