Posts by Filippo

Re: I give it 30 years...

It doesn't require much fuel to get in a state where you'll eventually fall. It does take quite a lot of fuel to fall quickly from a stable orbit.

Re: You can even sign in...

I have a Google account, for Google stuff. I never use it to sign in to anything else. I have an account for each service, and a password manager. It's not hard at all.

Re: If you can see it from the ground it's avoidable.

Yeah, they are flying at mach stupid. And most other things in orbit don't have any armor, because of weight. The quote I've always heard is that flecks of paint are still dangerous, and I see no reason to disbelieve it.

All of that said, removing the big stuff is a good idea. A big piece of junk is just a million small pieces of junk that haven't been hit yet. Taking it down while it's still in one piece is way, way easier.

Re: Good for Huawei

Basically, by the time you have found solutions for all the small and not-so-small problems that arise out of using large-scale IPv4+NAT, you have created something that's at least as messy as IPv6. And that's the answer to the OP.

Re: Interesting tech

I'm going to take a stab at answering, from the towering height of knowing nothing at all.

>So, what's the security on this ? Are you going to be broadcasting your brainwaves for anyone to read them ?

I'm fairly sure these are wired. I guess the ends of the printed wires are going to be taped to an actual wire, which goes into a recorder. I can't see wanting to add radio noise to an already-noisy signal.

>I have to wash my hair every morning. What's the impact of that going to be ?

You can't wash your hair while wearing these. But, good news! It's okay, because you have to shave nearly-bald anyway. Also, it's just for 24 hours. The exam, that is. Fixing the shaved-nearly-bald thing is going to take longer.

>And how long do you need to read an EEG anyway ?

I don't know, but if this works like many other long-duration medical scans, the point is not so much how long it takes to read the EEG, but rather the ability to capture unpredictable and short-lasting events, that could happen at any moment during the day. For example, there are heart exams that involve taking your pulse and pressure continuously for a day, even though those are trivial to measure at any given time.

Re: I had one user...

Normally, yes, for as an ordained tech-priest of the Machine-God, I can perform a techsorcism.

For a fee.

Praise the Omnissiah and pass the invoice.

Re: "Possibly, AI is not the big bonus that everyone's thinking"

>I'm thinking of the early days of computing and punched card machines.

But there are critical differences. Early computing devices had lots of problems, but they were well-understood problems (by trained personell). They sometimes did not have any technical solution, but they always had a clear theorical solution. When you're in that situation, the technical solutions usually show up soon enough, and they did.

The main problems with LLMs are not well-understood even by specialists, and they don't have any clear theorical solution. When you're in that situation, you'll be stuck there until someone comes up with something new, which might happen at any time or never.

Re: Asimov only wrote down some words to sell some books

Agree. I feel that the whole point of Asimov's robot stories is that you cannot have true intelligence AND reliable behavior boundaries. Those two requirements are directly at odds, even though this is non-obvious.

This is something the people who try to design LLM railguards should think about.

Re: What's the definition of a "political ad"?

I've just read the article and I don't know the law either. That said, from what I understand, a "political ad" for this law is anything that relates to an upcoming election or referendum. So a pro-abortion ad would be covered by the law if there was a referendum on abortion coming up.

I've no idea whether that would extend to pro-abortion ads if there was no referendum, but there was an election where abortion was considered a major issue. That ambiguity could actually be resolved in the law for what I know, but if it isn't it might be one of Google's reasons.

I have the feeling that a big part of Google's problem is that, even if there was no ambiguity, there is no automated way to figure out whether an ad is covered or not. There is no central database of all elections and referendums. Even if there was, ads do not come pre-tagged, and so-called AIs are not reliable enough to identify something that could land you in court if misidentified.

Google's business model relies strictly on full automation; they can't afford having a human vet ads.

All things considered, I'm fine with the end result. Most political ads give zero or negative information anyway.

Re: What is the point...

>Copyright holders can't make it illegal for you to read and digest a book, or even to memorise it down to the last comma. Why should they be allowed to stop an LLM doing the same?

Nobody is suing a LLM.

All of the lawsuits are against the LLM's developers.

LLM developers are not reading and memorizing books. They are feeding books to a computer algorithm, running it, and commercially exploiting the output.

During this process, let me reiterate, they do not read books, nor do they take any action that remotely looks like reading books. They mostly don't even know which books they are processing,

The LLM does do something which looks like reading books, but that is irrelevant because, again let me reiterate, nobody is suing the LLM.

On the other hand, doing computer processing of something under license and then selling the output is very unlikely to be covered by the license, and if it were, it would be in the list of things you are explicitly forbidden from doing.

I really can't put it any simpler than this.

Also, since this is something that comes up, just because something is freely available to read and download, it is not in the public domain. It just means that there aren't technical restrictions, which makes the legal restrictions hard to enforce. But they still exist.

Re: Cross laminated timber, eh?

>Cutting down trees that eat CO2 to reduce CO2 - WTF???

"Trees eat CO2" is a gross oversimplification. It's true enough for slogans and for kids, but not for serious talk. I think that our current tendency to get bored after reading more than four words in a row is a pretty large contributor to society's ills. Unfortunately, reality has no obligation to be describable by snappy soundbites alone.

In this specific case, cutting trees down to build something more durable than those trees (and no, not all trees live for centuries), and then replanting them, will reduce atmospheric CO2 more than leaving them alone. High school education and about thirty seconds of thinking should be enough to understand why.

Re: China is setting up the BRICS collaboration

Spot on. The article's author says "And true sustainable parity will depend on access to the global markets with competitive cutting-edge products. There's no sign of that." and showing he's missing this point entirely.

India + Russia + China + chunks of SE Asia, Africa and South America is a global market. And saying there's no sign of "competitive cutting-edge products" is myopic. There are all the signs, just not for next month, not next fiscal term, maybe not for a few years - but, guess what, there's life beyond the next few years, and that life looks a lot like Chinese silicon reaching the cutting-edge.

As for why US politicians went down this route - I'd say ignorance (history shows pretty clearly how this sort of thing works), short-termism (look tough now, it'll be someone else's problem later), but there's a lot of racism in there. I read "China can never reach parity with Intel/AMD" and I don't know how else to interpret it. Of course they can, they just needed a strong motivation, and we just gave it to them.

That's actually reassuring in some twisted way, thanks.

>They'll be even more profit when all the music played is generated.

There will be more profit - until the users figure out that, since they are getting generated crap anyway, they can generate it themselves, either on their own hardware or for a pittance. We'll see what happens to profit then.

Re: a step towards the realization of concepts confined to science fiction, such as The Matrix

>see: Does Time Really Slow Down during a Frightening Event?

Man, why don't I get to throw people off a cliff for a living?

Re: Unregulated Nuclear Generation ?

They might have to, if some future administration decides they want to show how tough they are on "big government" by neutering regulatory agencies and damn the consequences.

I don't know about the best use, but there are many uses that are way better than this. There's a list here: https://www.blm.gov/programs/energy-and-minerals/helium/about-helium

Re: Simplest solution

Code is a resource, but creating code is a service. A very scarce service. Getting paid for that is not evil. Attempting to get to use a service against the will of the service provider is usually evil.

Also, code is a resource to which the creator has full moral rights. That means that I, and I alone, get to decide whether adding artificial scarcity to code I created is evil or not. You do not get to determine that, and I do not have to justify my decision.

Philosophy, I know, but since you brought the "evil" word into the conversation, you get the philosophy. It's not always intuitive, but sometimes truth is complex and we just have to deal with that.

Re: Can someone please explain

It's... somewhat better. I wouldn't say it's "much" better.

The difference is that if you dig up a fossil and burn it, you are adding carbon to the carbon cycle that was not there before. Not on a time scale that makes any difference to anything. It won't go back into the ground.

If you burn wood, you are adding carbon to the atmosphere, but it's all carbon that was in the cycle already anyway. The plant would eventually have died and decomposed, on a time scale that varies a lot but is generally in the order of decades.

It's still bad, but it's a different kind of bad. By burning fossils, you are inflicting permanent damage; by burning wood you are inflicting temporary damage. It's a pretty long "temporary" and it has exceptions, but still.

That's only from the CO2 point of view, though.

There's the we-need-land-to-grow-food point of view, which, as you rightfully note, is a whole lot more urgent and IMHO should be sufficient to relegate biofuels to the dustbin of bad ideas.

Re: i guess she's saying that law enforcement is pathetic

Law enforcement is usually best employed as a last-tier solution to any given problem. It's good at dealing with point events, but it's extremely expensive.

Because of that, it's smart to deploy other measures to reduce issues in a statistical sense. Use incentives and disincentives to reduce motives. That way, the number of events that has to be dealt with by law enforcement can be minimized.

That is efficient. For some problems, it's the only way to realistically keep them under control. That is typical of crimes where victims and perpetrators are in collusion, such as contraband, drugs and, yes, ransomware.

Expecting to solve all problems via law enforcement alone is like allowing everyone to just dump trash in the street, and then having the street cleaning staff deal with it. Sure, it works in theory, if you think it's sane to spend the entire town budget on the cleaning staff.

Re: That's not a Stapler!

With wireless charging, if something goes wrong, it could be a microwave oven.

Re: Not so sure about this....

That feels like one of those evil genie stories where the character tries really hard to craft a wish that won't screw him up. The reader knows that the only correct course of action is to not use the genie at all, but the character never does.

Re: ‘Bout time someone bit back

If this resulted in Epic being forced to allow third-parties to sell Fortnite skins outside the store, I'd have a hard time stopping laughing.

Re: C

The technical reasons for cybersecurity being a mess are very nearly irrelevant, when compared to the economical reasons.

You can bet that if cybersecurity incidents routinely resulted in lawsuits and/or criminal charges against the vendor, buffer overruns would be made extinct PDQ some way or another.

Re: "Programming should be easier"

That is exactly how you end up, after a few years of doing this, with unmaintainable messes that are easier to throw away and redo from scratch than to fix. That's only efficient if all you care about is next release and next fiscal term. If the objective is to make a product that will keep making lots of money for decades while being cheap to maintain, it's an extremely inefficient approach.

Good points. Unfortunately, there's a good number of people who, when encountering the word "nuclear", just stop reading and start raging. There's also a good number of people who exhibit a similar behavior on reading the word "Microsoft". Having both words in the same article is pretty much a guarantee that the comments section will be useless.

When I'm feeling optimist, I hope that first big tech builds a crapload of carbon-neutral power generation for AI, and then the AI bubble bursts for good, leaving said generation available for useful stuff.

Re: Conspiracy

>Not remotely the same thing.

How so? If the premise is that a hostile actor could hide a remote-controlled explosive charge into a car, what does it matter whether it's an EV or not? You'll be killed by the explosive, not by whatever the car runs on.

Absolutely. But it's also difficult to imagine an insurrection without a cult of personality. Few people would storm Congress for Biden or Harris.