Poor poor persecuted white male. *pats head* there there, it'll be alright. Show us on the doll where she touched you and tell use how much you liked it.

I bet you leave all your front door unlocked and open at home since someone could just break a window and therefore your house will never be safe.

Yes and so should you. Why should you be forced to pay extra to someone multiple times through every piece of software you use, because they did some math?

No it begs no such question. But to answer your independent question that wasn't begged, no there's no patent encumberment in JPEGXL, it's defined as an IEC standard, and has a reference implementation free for use under the Apache 2.0 license.

Unused, and no use case for it, no pathway predicting it would become popular, and no reason to develop it further.
That changed when the PDF/A decided to adopt JPEGXL officially. Now PDF readers need to implement JPEGXL and both Chrome and Firefox include PDF readers in the browser. Their hand was forced.

Nothing has turned into a religion. Programming rust is safer than not programming in rust due to the nature of the language and how it needs to be used to create the same class of bugs. That's not religion, that's a simple verifiable fact. No one said it's "safe", they said "safer". There's a previous implementation in a language without memory safety to compare it to.

No one said that. No one said it's "safe". They said it's "safer". Memory-safety makes the whole codebase safer than not having memory safety.

Now go look up what a "comparative adjective" is so you don't say something stupid again in the future.

So you're saying if the question isn't defined before hand the result ceases existing? Not even Schrödinger could justify a comment that stupid.

We have a long history of exploiting decoders. Image based malware exists, has for decades. Hell back in 2004 the Windows GDI+ library had a arbitrary code execution attack trigged by simply opening a Office document with a specific embedded JPEG image, or using Office to attempt to insert a specific JPEG image into the document. I don't think IE was affected so it wasn't a remote execution from what I recall. That's just one example. Exploiting of decoders is a thing.

The attacks often exploit standard software bugs, e.g. memory unsafe code, buffer overflows etc. In this case they are talking "safer" because they are using the Rust based library which prevents a subset of the potential bugs creating security issues.

Don't confuse your ignorance with marketing trash.

That's cool, I'm currently at an airport on a mobile connection. How long of an ethernet cable does my NAS need to reach the next continent?

Not all use of AI is slop. AI is a tool. There's no reason it can't be sensibly incorporated into a workflow. While no doubt creating a game from a prompt is just slop, the reality is there is a practical place for including tools such as this in a workflow, such as using it to create starting point.

I recently came across this at work during a 1 week workshop we had. All the presenters (including myself) used AI extensively to create our presentations (because powerpoint fucking sucks), but only some of them looked like slop. The ones which looked like slop were the ones where AI did the finishing: i.e. people made presentations and asked AI to make it pretty. The rest of us used AI to structure the presentation and then manually tweaked it afterwards and they looked unique and individual.

There is a place for this. Use an AI prompt to come up with a base game, and then put AI (Actually Intelligent) people in place to turn that base structure into something creative.

It's not as clear as you think. It's not sufficient to write and tweak prompts. Look to Jason Allen v. US Copyright Office https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.artnews.com%2Fart-ne... Jason Allen attempted to copyright art that he created with Midjourney. Despite going through hundreds of prompts and using photoshop to edit the result *he* (not the computer) was denied copyright by the US Copyright office.

I think the court case is still pending.

The keyboard seems fine. I think I'll be able to type prompts at a sufficient WPM.

Benchmarking CPU/GPUs targeted at laptops generically is fucking stupid. The performance of each device is almost exclusively determined by the thermal and power design. You are using benchmarks wildly incorrectly. Please stop.

Now maybe you're right and the performance numbers are as you suggest, but if you are then it's by fluke, not because you're doing anything sensible. Wait until the device is released, tested, and benchmarked in its actual configuration before you draw any conclusions.

And no portability? If you're in the market for a desktop then this product isn't targeted at you. It's okay, not everything is made for Fly Swatter (30498). I know you may not like to hear this but there are ... other people out there. At least for now. God knows the way the world is going you may be the last survivor in your basement.

Literally every desktop computer thermally throttles as well. It's a question of time. CPU / GPU design has dynamically adjusted performance to thermal envelopes for years now. It's why watercooling your system often nets you performance boosts even without changing a single setting.