Especially since you know the AI tools are going to be enshittified in short order. That ChatGPT query is free today, tomorrow once you're hooked and can't live without it, it's going to be a dollar query. 50 cents if you can wait for it to be answered during "off peak" hours.

Because that's the business model that's likely to win out. Get kids hooked on AI then get to bill everyone $50/month to use it because they've gotten so reliant on it that they can't live without it.

The whole AI causing shortages is to ensure that you don't get a chance to run local models yourself without paying for the privilege.

The bigger problem is that OpenAI didn't actually log what the agents did so we only know this because the victims spoke up. It wasn't just Hugging Face, but various oldschool wikis in Germany, The University of Toronto and others who were accessed.

I don't know how they sandboxed the AIs but it didn't seem to do a good job if it iddn't log what sites it was using and Hugging Face was the one who detected the hack.

I saw part of the testimony. This all comes down to human over-aggressivenessâ"not even mistakes.

The detective used Flock to look for all Dodge Durangos within a radius of the accident. They located her at her apartment. So far, so good. But her car is a different color than what eyewitnesses said. Other details did not match up with eyewitness accounts. There was NO damage to her vehicle, but they arrested her anyway. The detective doubled down and the police continued there prosecution of her.

I'm no fan of Flock, but in this case it's simply bad police.

And the same detective could've queried the DMV and asked for all Dodge Durango vehicles registered in the state. They have that access by default. Sure, they would get a lot of hits, but then they could narrow it down, by partial plate, color, age, etc.

And likely she never would've even been a suspect.

No, Flock may not be completely guilty, but it's not innocent either because just passing by should never be a reason to be arrested. The only way for Flock to win is if its camera captured the accident because it happened to be right in front of it.

It's poor policing, but it's Flock-enabled poor policing.

MW Propagation is a lot of fun. You can get stations that are thousands of miles away. Too bad younger generations haven't bothered to experience it.

It's a huge PITA for AM stations. Many AM stations actually have licenses that say they can broadcast from sunrise to sunset and must shut down for the night. Others have "reduced power operations" where they must reduce their transmit power overnight.

And others are lucky to have the full license that lets them go 24/7 at full power. The regulatory bodies have to consider propagation when dealing with AM radio licenses. When an AM radio station goes off the air it presents more headaches because the other stations nearby might have to juggle their power settings.

But AM radio being in the HF radio band means you only need a few stations to cover an entire country. In an emergency, you cannot tell how many transmitters you have available so the AM network means you still have a way to communicate even if stations so down. And some AM stations do not operate at full power as full power operation is reserved for emergencies.

In a scenario where a major city is impacted, nearby AM radio stations can often have coverage into that major city despite being far away (and thus not experience the issues) making it the ideal communications medium.

Plus, it also means the EV makers can't cheap out on radio shielding. ICE vehicles have huge issues as well with EMI causing radio issues, but they have had years to shield the radio from interference caused by the engine. ("Alternator whine" is a common one). The only reason EVs want to ditch AM radio is because their electronics interfere with AM reception and they want to save the $5 of metal to shield components.

Gone With The Wind made all that money with $6 tickets, adjusted for inflation. Average ticket prices in 2019 were $9. And there were a third as many people in the US at that time.

Not really. When GWTW came out, movies were basically the only entertainment option out there as we would recognize it. TV was still experimental and radio was the only thing you would do at home. Beyond that you have books and hobbies.

Spending a Saturday at the movie theatre (the "nickelodeon" because it was 5 cents for entry) was a popular pastime - you'd get a few cartoons, a newsreel, then a couple of movies. This was popular as it was a way to catch up on the news and watch the latest movies that came into town.

Gone With The Wind was extremely popular because it was a good movie for the time so it would often be a feature that showed week after week because people kept demanding it.

You also have to realize the "new movies" were whatever pulp Hollywood came out with because they knew they were just going for the nickle movies.

These days, we have far more entertainment options so getting a dollar at the movie is actually much harder. After all, you can watch a movie in the privacy of your own home nowadays, something you couldn't with GWTW. If you wanted to see it again, you had to pay for another viewing.

EDK is the EFI Developer Kit. There was one released by Intel which wasn't open-source, but there have been several EDKs that are open-source. But they are for developing UEFI applications that run in the UEFI environment. You don't use the EDK as a replacement for UEFI.

LIbreBoot/CoreBoot and now GNU Boot are a BIOS/UEFI replacement. And they're actually all derived from from CoreBoot (which started as LinuxBoot). It basically is a BIOS/UEFI replacement that booted into Linux and served as the boot environment.

CoreBoot evolved from it but it has binary blobs, which is why the GNU folks created LibreBoot. Not surprisingly the compatibility list of LibreBoot is far smaller than CoreBoot, and CoreBoot is pretty small already. Remember this has to initialize the hardware the provide the tables so the OS can perform certain operations like sleep and suspend by telling the OS what registers it needs to access.

LibreBoot has a sordid history because earlier this year or last year, the developer of CoreBoot got fed up with LibreBoot being years out of date, so the CoreBoot dev basically pushed an updated version of LibreBoot (minus the binary blobs). Apparently the LibreBoot dev didn't like that, which is why it's GNU Boot. It was a pretty messy thing because the CoreBoot devs wanted to update what was a completely outdated fork.

Riot isn't identifying a Ryzen by staring at TCP packets. Vanguard runs on the PC, where it can collect hardware information or request TPM attestation and send the result to Riot. CPUID identifies a processor model, not an individual chip; a CPU-bound firmware TPM is a more plausible explanation for a ban following a used CPU.

TPM wouldn't work because you'd have to transplant it from one motherboard to another. The CPUID instruction gives a lot of information (including serial number) if you ask it the right way. It's how the CPU-Z application gets all that information - there's tons of information available via CPUID on a modern processor depending on what you load EAX/EBX can call it.

How will you get the ROM once everything is "digital"?

Hack the PS5.

There's a new jailbreak on PS5 that handles basically up to firmware 15.5. Use it to get access to the PS5 filesystem and grab your files from there.

I'm sure there will be other ways that probably involve simulating the PS5 and downloading the files directly.

I'm not worried about the idea of personalized pricing. It just doesn't seem feasible. Implementation hurdles are abundant, the likelihood of complaints in-store is very high, and for not that much benefit.

Personalized pricing in store will be a nightmare for Walmart. Think of several situations that are extremely common. Like what if two people are looking at an item? When I shop at Walmart, there are inevitably other people around and they're looking at the same stuff I'm looking at. The price tag can only show one price - so whose price is it going to show?

The second problem is shoppers in general. There are people who take time to shop - they can take an hour or more shopping for groceries and stuff. You're going to have to remember what price you shown someone so when they check out you use the "right" price. Also, if you've done grocery shopping, you inevitably run into people who complain about every price as it's rung up. Imagine the fun to be had if the price changes and they're constantly checking to make sure the register shows the correct price.

And if you're going to do dynamic pricing, people are going to snap photos of the price tag and make sure they're being charged what the tag says. It's going to clog up the checkout lines because people are going to verify every price rings in properly. This just makes the lines way longer and way slower, and likely having to have more cashiers on checkout will negate any extra profits to be made if checking out takes way longer as everyone argues prices.

And yes, people will argue about things like 10 cents or less.

Taction probably is a troll, but ever since Apple sued Samsung for round corners design patent infringement, I figure that made Apple a patent troll too, and every patent fight they lose is karma. I'm on Taction's side for that reason alone.

And let the ignorance of the rounded corners patent show.

The rounded corners patent was a design patent. The Taction patent is a utility patent. If you don't know the difference, well, a utility patent is on something the thing does. A design patent is on something it looks like.

They are patents in that everything is AND - you must have something AND something AND something AND something ... If you lack one of those things, you are much more likely to win a patent lawsuit.

With that in mind, the Apple patent is simple - rounded corners, a tray with icons and a app selector that scrolls left and right. Samsung had that to a T. They literally copied the UI. Why didn't Apple go after Google? Because Google didn't. They had rounded corners, an app drawer at the bottom, BUT, not an app selector - because Android had widgets you could put up.on the screen. They had a separate app drawer that showed all the apps. Those two differences meant "rounded corners" patent wasn't valid on Android in general.

It's just like how the McDonald's hot coffee lawsuit was much more than hot coffee and once you understand that, the lawsuit made much more sense. (McDonald's purposely sold overly hot coffee to save money on coffee grounds as it let them extract more coffee from them. They were told time and time again it was too hot to serve but they persisted. When it happened, the victim asked for a mere $30,000 for medical bills. But that was rejected, despite the fact that McDonald's paid out medical expenses for previous occurrences. And while she won millions, it was reduced on appeal to under a million dollars).

It's much more than rounded corners. It's literally Samsung redesigning the UI to match the iPhone in functionality. When Samsung lost, they basically added widgets to the screen which was enough to work around the patent. Eventually they decided to make TouchWiz their own.

Lots of keyboards ditched more useful keys to fit a copilot key,

Actually, the key the Copilot key took over was the Context Menu button on most keyboards. If you pushed it, you got the equivalent of the right-click menu. Unlike the Windows key, it likely never got much use, so it remained a mostly useless key.

It's so useless many keyboards chose to omit the context key or put other keys there. Some replaced it with another Windows key, others used custom function keys.

Their reasoning sounds plausible only if read from a "business weasel" perspective. If there's a market for RPis with more lower quality RAM, nobody is "undercutting" RPF; they're selling the products they want to, at the price they want to, and choosing not to offer big/slow/cheap OEM RAM configurations. And even if someone could undercut them on the exact same components - So what??? RPF may be a non-profit, but I am not!

The problem isn't resellers undercutting RPi. It's someone buying a 1GB RPi, doing the mod with 8GB of poor quality RAM, then selling it as an 8GB RPi at a modest discount.

The problem is that you buy the board and because the RAM is bad, you perform a return and find out it was a counterfeit RPi board that doesn't work for you.

Despite the AI issue, you can find RAM. Often poor quality recycled RAM with errors are stupidly cheap (and used in many products as-is). The board can boot with bad RAM, but run stable it might not, and you'd be more likely to blame RPi than on the reseller you bought the dodgy unit from

The fact that RPi took this long to do it would mean they're actually seeing a bunch of returns with the RAM swapped out.

It could also be that people are buying the 8GB units, swapping them with 1GB units and selling them as fake 8GB units, like those $20 "1TB" USB sticks you can find easily on Amazon.

Changing the RAM on an RPi isn't a trivial procedure - I believe it's PoP RAM so it's not something a hobbyist would easily do without a lot of specialized equipment and experience in being able to separate PoP RAM from the base CPU and reballing and such. So hobbyist wise, they aren't likely to be ones affected since few would have the expertise or equipment. But those trying to sell counterfeits do.

but what do we call the person who still supports Trump, but *isn't* stupid? They're smart people, but they support this craziness. Is it fascist? Is it evil? What's the correct term for this?

Try "ignorant". They can be smart people, but ignorant in refusing to see the truth.

It's why a common phrase now is "Have the day you voted for". When you see these Trump supporters complain about high gas or diesel prices, you say "you voted for it, what are you complaining about?" All the information was there for you to analyze, but you chose to be ignorant on purpose.

Ignorance is a form of stupidity. Have the day you voted for.

People have a reasonable expectation of privacy. Piecing together information from the thousand or so cameras that see you everyday is a violation of the 4th Amendment if sold to the police or otherwise is acting on the direction of the police. They need a warrant and Chatrie v. United States is only the beginning of how privacy violating tech companies are going to get slapped down.

Even that's not as creepy as Flock giving access to anyone who installs it access to the entire network.

It's the network effect - the whole is greater than the sum of its pieces - multiplied. A LEO in some rural part of a flyover state, because their town installed Flock, now has access to Flock cameras in Houston or LA or New York.

In other words, the potential for abuse is too tempting. Even if it was for legitimate law enforcement uses, if you required a warrant to get information on a car, you could ask your buddy to do the lookup for you located across the country where no warrant is required and get the same information.

It's funny because the other ALPR systems out there (e.g., Axon, etc) restricted users to the cameras they installed. So you could stalk your ex if they lived in the same town or city, but if they moved, that's it. Once they left city limits you can't track them anymore without cooperation from another agency.

It's not that ALPRs are bad - you probably run into a ton of them as they're everywhere. They handle your toll charges if you use a toll road or bridge without having to stop and pay at a booth. They handle your parking (did you know chalking a tire can be considered a 4th Amendment violation? Well, that just means cities switch to ALPRs to handle their parking enforcement).

And when Flock started, it started as a way for private individuals to monitor their traffic - private companies would install it to their parking lots, HOAs on their roads, etc. But it was again limited to just those uses - the HOA could only get information about its cameras.

The question is ultimately if you would have been better served with doing your own investigation rather than relying on AI, did you retain the knowledge, and was the information complete/accurate.

I use AI, and it helps me navigate a code base far faster than without. Like if I need to know if a function is used, the AI could tell me in a minute, or I could be using grep on the code base for a half hour.

I know to not trust it and to verify the output. It crawled deep into a complex submodule I didn't understand and helped me devise a fix. I didn't check that fix in as it had a few errors code-wise that would have become a maintenance nightmare and a lingering bug if someone added a member variable to a struct.

Now, I could've gone through the code myself, and probably taken a few weeks to properly understand it. But AI did it, and I had it produce a few documents on the design of the system. Those documents let me understand what's going on so I can see how it's doing things and how I can actually see the code better. Granted, with a few false starts.

I would say the AI helped me - mostly because I was told to treat it as a co-worker than as an expert to ask to do something. So I would guide the AI into what needed to be done, and then leave the final details for the AI to do it. Then I would review what the AI did

My AI token spend isn't huge compared to other users, but that's because I'm judicious with how I use it. It spends more time doing cheap things before coming out with a result that's relatively succinct and correct so it's not wasted on expensive output tokens.

Everything I used AI for, I could've done it myself. It just means I have an assistant who saved me likely a month of work trying to understand the code and another month devising a fix and I managed to get it done is about 2 weeks and in a way I could understand what it did, and the code reviews of the changes make a lot of sense.

I still remain a skeptic and find these highly guided uses make me more efficient while letting me not get too bogged down in the details. I constantly question whether I could do it myself or if I'm relying on the AI and make sure the AI output makes sense. (I don't trust it to commit code to git, for example, so I have it make all the changes it wants, then I test them and review them before submitting).