Yeah, it's not even worth considering for something like 15-20kg. A full pallet in this case is 464kg :) Years worth of supply for me.

The current "AI" is a predictive engine.

And *you* are a predictive engine as well; prediction is where the error metric for learning comes from. (I removed the word "search" from both because neither work by "search". Neither you nor LLMs are databases)

It looks at something and analyzes what it thinks the result should be.

And that's not AI why?

AI is, and has always been, the field of tasks that are traditionally hard for computers but easy for humans. There is no question that these are a massive leap forward in AI, as it has always been defined.

It is absolutely crazy that we are all very very soon going to lose access to electricity

Calm down. Total AI power consumption (all forms of AL, both training and inference) for 2025 will be in the ballpark of 50-60TWh. Video gaming consumes about 350TWh/year, and growing. The world consumes ~25000 TWh/yr in electricity. And electricity is only 1/5th of global energy consumption.

AI datacentres are certainly a big deal to the local grid where they're located - in the same way that any major industry is a big deal where it's located. But "big at a local scale" is not the same thing as "big at a global scale." Just across the fjord from me there's an aluminum smelter that uses half a gigawatt of power. Such is industry.

That "ruler study" was ancient. It's mentioned in peer review at least as early as 2018, and might be even older.

Believe it or not, people in the field are familiar with these sorts of things that you just read about.

Most of these new AI tools have gained their new levels of performance by incorporating Transformers in some form or another, in part or in whole. Transformers is the backend of LLMs.

Even in cases where Transformers isn't used these days, often it's imitated. For example, the top leaderboards in vision models are a mix of ViTs (Vision Transformers) and hybrids (CNN + transformers), but there are still some "pure CNNs" that are high up. But the best performing "pure CNNs" these days use techniques modeled after what Transformers is doing, e.g. filtering data with an equivalent of attention and the like.

The simple fact is that what enabled LLMs is enabling most of this other stuff too.

Jevons Intensifies

If you can use your phone to control everything, then the only device you need with you is your phone.

And when you lose your phone or it's stolen? Or if it somehow gets bricked after an upate? Or gets compromised?

There's this thing called single point of failuare.

Agreed. If someone wants a movie or music or software, they can request it from the owners and can only use the item they were specifically given permission to use. Otherwise, they have to pay for it.

> Meanwhile, H.264 has dedicated hardware decoders in world+dog devices, including ancient ones.

Ancient ones, yes, but most devices sold in the past five years have AV1 *decode* support.

Hardware with AV1 *encode* is still pretty rare but a fair number of up-market chips from the past few years have it.

What we mostly care about here is the $20 amtel or mediatek devices sold today, and those are fine.

Netflix can support the older devices with H.264 as long as it makes more sense to pay the patent license fees than to drop support for old devices.

It won't be long before there are no devices that the manufacturer still supports that can't decode AV1 in hardware. Not that most end-users even know their device went EOL and now a potential liability.

Given that Netflix has native apps on most of these systems it should be straightforward to serve the non-patented stream to any device that can play it well.

> They don't do backups at those outfits?

We really need Federal government backups to be centralized at the National Archives.

Both so one expert team can make sure it's done right, instead of hundreds of teams with questionable experience and track records attempting to do it right.

And /also/ so when one agency goes, "whoopise, I guess we deleted the evidence of our crimes!" there is recourse.

Right now, the prosecutor just goes, "shucks, I guess we don't have a case then. Better fire some leaf-node IT contractor."

The only reason they deleted the databases was because they were let go by Trump. Most likely because they were Biden hires.

How many other hackers who have served time are subsequently hired by the government? Should we complain about that as well? We know of at least one who ran a site which offered services to hackers, and he was given the keys to the kingdom. How much private data do you think he exfiltrated?

The real issue is, once again, a private company didn't follow basic security procedures and let these two have access while they were being fired. Standard practice is to disable all access and escort from building when being fired. Neither was done in this case.

I've done my first test of buying a whole pallet of filament straight from a Chinese manufacturer. It's a risk - it could be all junk - but if it's usable, the price advantage is insane. Like $3/kg for PETG at the factory gate (like $5/kg after sea freight and our 24% VAT). Versus local stores which sell for like $30/kg.

I'd love to see someone try to 3d print with a filament that melts at 162K. Where do you even buy xenon filament? ;)

Yeah, if you had injected moulded PLA, it would have been just as terrible

Early on, I was overdoing chamber heating, and later discovered that was part of my problem. A blanket and a duvet can get a P1S's chamber over 70C. But if you do that, in my experience, like half an hour or so into the print you'll get heat creep problems and the filament will split & the extruder will just dance around in the air as though it were clogged (though maybe my filament was just garbage... it certainly was *wound* terribly). I ended up using a meat thermometer stuck in through one of the holes to measure temperatures, and then I'dadjust the positioning of one small blanket over the chamber to try to keep it in the mid to upper 50s, and was able to finish big prints that way.

But yeah, whatever means you use, you need some sort of raft and very strong reinforcements.