If this was a poorly drawn Paint illustration it probably wouldn't have been as funny or recognizable. AI is fast, easy, and good enough. This is just a meme most people will forget about in literally 1 minute, IMO there is nothing wrong with using AI like this
I've always wondered why there isn't VR optimized compression. Seems like each eye would mostly have similar or shifted pixels than the other. Nice to see some work being done in this space
A lot of Windows bugs are memory corruption bugs. And those are often severe. Using Rust does actually prevent memory corruption. The rest of Windows is still probably insecure, but any security improvement is good.
No, it's expensive to comply (at a massive scale), but easy to avoid. Just change the user agent. There's even a dedicated extension for bypassing Anubis.
Even then AI servers have plenty of compute, it realistically doesn’t cost much. Maybe like a thousandth of a cent per solve? They're spending billions on GPU power, they don't care.
I've been saying this since day 1 of Anubis but nobody wants to hear it.
Given the difference in server CPU/memory/storage/network/scale, I don't think it's possible to get any number with confidence. Maybe you could self host, but that wouldn't be representative of real email servers. Plus different email providers handle emails differently.
And cloud providers probably automatically scale with load, Gmail probably uses more power during work hours than after.
Also SSR is relatively new and email services are ancient, I'd be surprised if any used it. I'm not even sure if it's a good idea for email.
Plus it would probably vary with how many emails you have in your inbox...
I just don't think it's possible to get an actual number.
Networking is remarkably efficient, and so is decoding images, because processors do it so quickly. For more intensive tasks like video, hardware decoders make them efficient. All email is to the client computer is making a request for data, processing that into a list that can be displayed, and displaying it. I'm pretty sure just having the screen on is orders of magnitude more power hungry.
The point they're trying to make is that while Apple location collects anonymized data, the app itself could collect that data and send it off with a unique identifier. Like Strava except it's spyware. It has location and internet permissions, and the App Store data privacy thing is self reported and not necessarily factual. We can't know unless we dissect the app because it's not open source. Right now it seems they're not doing that, but nothing is stopping them.
Android's Fused Location Provider, even though it's provided by Google Play Services, works in the same way and is just as private as Apple's location API. They are both on device with anonymized data sent back. So IDK what the devs are talking about. But I am interested in their blog post.
Fused Location Provider data sent to Google is anonymized with a temporary device id anyway. They also can't tell if you needed location for this, maps, or Pokémon Go. But they do sometimes collect GPS, WiFi, and cell data. Not using it is more private but I don't think it's worth worrying about.
Nothing isn't a "major OEM" but I think they'd be interested in it. Maybe they'll be next.