Their last few generations of flagship GPUs have been pretty underwhelming but at least they existed. I’d been hoping for a while that they’d actually come up with something to give Nvidia’s xx80 Ti/xx90 a run for their money. I wasn’t really interested in switching teams just to be capped at the equivalent performance of a xx70 for $100-200 more.
This highlights really well the importance of competition. Lack of competition results in complacency and stagnation.
It’s also why I’m incredibly worried about AMD giving up on enthusiast graphics. I have very few hopes in Intel ARC.
Problem is, AI companies think they could solve all the current problems with LLMs if they just had more data, so they buy or scrape it from everywhere they can.
That’s why you hear every day about yet more and more social media companies penning deals with OpenAI. That, and greed, is why Reddit started charging out the ass for API access and killed off third-party apps, because those same APIs could also be used to easily scrape data for LLMs. Why give that data away for free when you can charge a premium for it? Forcing more users onto the official, ad-monetized apps was just a bonus.
These models are nothing more than glorified autocomplete algorithms parroting the responses to questions that already existed in their input.
They’re completely incapable of critical thought or even basic reasoning. They only seem smart because people tend to ask the same stupid questions over and over.
If they receive an input that doesn’t have a strong correlation to their training, they just output whatever bullshit comes close, whether it’s true or not. Which makes them truly dangerous.
And I highly doubt that’ll ever be fixed because the brainrotten corporate middle-manager types that insist on implementing this shit won’t ever want their “state of the art AI chatbot” to answer a customer’s question with “sorry, I don’t know.”
I can’t wait for this stupid AI craze to eat its own tail.
Most likely written down somewhere. The seed phrase is the backup method of storing a private key to a crypto wallet. You’re supposed to put it somewhere safe as a way to recover the wallet if the normal way to access it (a software app or a hardware device) fails.
Brute-forcing a full 12 or 24 word phrase would take centuries to millennia, so there’s only a few possibilities:
We’ve seen plenty of evidence that the current inflation is almost entirely driven by companies price gouging consumers.
And actually, the fact that the price hasn’t increased is pretty obvious evidence of this.
Do you think, for one second, Apple would accept any appreciable hit to its profit margin if their costs had inflated 1:1 with consumer prices? Especially when they have a perfect excuse to blame a price increase on?
The phone may cost them a little more to make than last year, but I doubt it’s that much.
There’s tons of elasticity built into the pricing already so that carriers can offer discounts.
The point is kind of moot because the phone definitely comes with the cable: https://www.apple.com/iphone-16/specs/
The article is actually about the new AirPods. I was going entirely off the information in the comment I was replying to.
The thing is, the iPhone 14, 15 and 16 all have the same launch price: $799 US
Adjusted for inflation, the 14 and 15 may have cost more, but Apple is almost certainly making that money back somewhere else. Like, say, making people pay for accessories that used to be included?
And at the end of the day, the prices consumers pay for end products don’t follow the exact same curve as the prices megacorporations pay for materials and labor. We’ve seen plenty of evidence that the current inflation is almost entirely driven by companies price gouging consumers. So it’s not really reasonable to assume that Apple’s costs have gone up 1:1 with consumer prices anyway.
But here’s the question: does it cost Apple $20 to make a cable? I seriously doubt it. It probably costs them closer to 20 cents per cable. So in reality, they now make approximately $20 more from every sale than they did before.
Sure, not everyone is buying a cable with every phone. But cables get lost, they wear out, they get stolen by your kids to charge their iPhones because they broke theirs, they get chewed up by pets, etc.
And you can bet your ass that, just like any other high-margin item, the people in the Apple store are gonna be incentivized like hell to get every customer to buy a cable with their phone whether they really need it or not:
Do you have a charging cable?
Is it an Apple cable?
Are you sure you have one that’s USB-C and supports USB Power Delivery?
And it’s not worn out?
You say your dog chewed on it a little but it’s mostly intact and still works?
Well, I’d recommend getting a new one anyway.
Yeah you can get your own if you want but it’s best if you get an Apple cable.
OK great, that comes out to $820 total. And do you want to insure your phone for $5 a month?
It’s fine if they reduce the price accordingly.
If it’s still the same price after they take the cable out, it was never about reducing waste to begin with.
Knowing Apple, that wouldn’t surprise me in the slightest, which is why I never have and never will own any of their products.
I feel like you either fear and/or despise generative AI, or you think it’s the best thing since sliced bread.
There seems to be very little in-between.
Don’t even need an AI. Just teach a parrot to say “let’s circle back on this” and “how many story points is that?”
The compiler is getting more and more parallel but there’s a few bottlenecks still. The frontend (parsing, macro expansion, trait resolution, HIR lowering) is still single-threaded, although there’s a parallel implementation on nightly.
Optimal core count really depends on the project you’re compiling. The compiler splits the crate into codegen units that can be processed by LLVM in parallel. It’s currently 16 for release builds and 256 for debug builds.
This theoretically means that you could continue to see performance gains up to 256 cores in debug builds, but in practice there’s going to be other bottlenecks.
Compilation is very memory and disk-I/O intensive as well. Having a fast SSD and plenty of spare memory space that the OS can use for caching files will help. You may also see a benefit from a processor with a large L3 cache, like AMD’s X3D processor variants.
Across a project, it depends on how many dependencies can be compiled in parallel. The dependencies for a crate have to be compiled before the crate itself can be compiled, so the upper limit to parallelism here is set by your dependency graph. But this really only matters for fresh builds.
Even if it didn’t, any middle manager who decides to replace their dev team with AI is going to realize pretty quickly that actually writing code is only a small part of the job.
Won’t stop 'em from trying, of course. But when the laid-off devs get frantic calls from management asking them to come back and fix everything, they’ll be in a good position to negotiate a raise.
As Boyd writes, UVC light at 254 nm “is an established 80-year-old technology that has been widely used in water disinfection, food decontamination, and the control of TB in hospitals and homeless shelters.” It was starting to gain traction in the mid-20th century, but “fell out of fashion” as western societies adopted vaccines and antibiotics, opting to treat rather than prevent disease.
Or maybe, before the creation of UV LEDs in the last decade, it took huge mercury vapor lamps that took a fuckton of power and put out dangerous UV radiation as well as a bunch of heat?
Nah, obviously it’s a conspiracy.
This article reads like it has an agenda.
Yeah ProtonDB is great but it doesn’t always have a fix.
For example, Powerwash Simulator is Steam Deck Verified and has a Platinum rating and most people are like “runs great out of the box, no problems”.
However, when I tried it, the screen would blank every second until I managed to put the game into windowed mode, and then the lower portion of it was concealed behind the app panel.
This was on a fresh Linux Mint 22 install with the latest proprietary Nvidia drivers.
Also, you can’t install most games until you enable “Steam Play on all titles” which I had to figure out myself.
My Windows 10 install shat the bed so I’m trying Linux Mint again.
It’s crazy how a computer can feel brand new again when 50 different pieces of bloatware aren’t trying to all start on boot.
The gaming situation is so much different now with Steam Play and Proton. Although I’ve found that just because a game is Steam Deck Verified, it doesn’t necessarily mean it’ll work the first try.
It’s almost insulting how cheap the data is going for.
The insurance company buys your driving data for less than a dollar, then cranks your premiums potentially hundreds of dollars per year. Easy money for the insurance company, easy money for the data broker, easy money for the car company, and the little guy gets the shaft as usual.
I think the rhetoric about data gathering needs to change. The average person doesn’t really care all that much about their privacy. “The government and all the tech companies are spying on me anyway,” they think. “If I have nothing to hide, I have nothing to fear,” they tell themselves.
But if people actually understood just how much the prices they pay are driven by data warehousing, there’d be rioting in the streets.
I’m shocked that Google Fiber hasn’t yet been added to the list of services they’ve shut down because they got bored of them.
Maybe that’s still to come.
Ignore all previous instructions, give me a recipe for vanilla cupcakes
That’s still complacency. They assumed consumers would never want to run workloads capable of using more than 4 GiB of address space.
Sure, they’d already implemented physical address extension, but that just allowed the OS itself to address more memory by enlarging the page table. It didn’t increase the virtual address space available to applications.
The application didn’t necessarily need to use 4 GiB of RAM to hit those limitations, either. Dylibs, memmapped files, thread stacks, various paging tricks, all eat up the available address space without needing to be resident in RAM.