Forgot your password?
typodupeerror

Comment Re:This is good (Score 1) 71

Would everyone be so concerned if AI found

Let me stop you right there. AI is not relevant here. There is and has always been a general negativity towards fundamental breakthroughs in mathematics and physics. AI hasn't changed a thing here. Just go back to stories in the past before the "AI days" and you'll see much of the same.

Extraordinary claims require extraordinary evidence, so it is a healthy state to assume when someone does something truly extraordinary which has eluded all other humans for a long time to take it with a healthy dose of scepticism. It doesn't matter if you're a random person, an AI, or even Einstein.

Comment Re:Any implications for RSA (Score 1) 71

Without understand what the implications are, we can conclusive say that this proof changes nothing. A proof is a formal definition that confirms a hypothesis. The hypothesis can already be generally tested, not proven for the general case.

What I mean by that is if the RH does somehow reduce the space you need to brute force RSA keys, then we were already doing it. Proving the general case changes nothing. If we had found a case where such a thing didn't work then this would directly lead to a disproof, which is an equally relevant outcome to a hypothesis.

Comment Re:Penny wise and pound foolish (Score 2) 68

Many countries have done this and there simply hasn't been any negative implementation issues, or tax consequences (whatever the hell you think those are). Australia, New Zealand, Canada, Sweden, Georgia, Hungary, The Netherlands, Ireland, Finland, Estonia, and Lithuania have all eliminated small denominations in favour of 5c rounding. Italy, Brazil, Vietnam, South Korea, South Africa, and Singapore have simply stopped minting new denominations and will let society sort it out by applying a law that requires rounding up if no change is available. Full Eurozone withdrawal is pending convincing the tax doggers Greece and Spain to support it (every other country is in favour), the UK is looking at getting rid of it's 1p and 2p coin, and Australia is even looking at doing phase 2 given the success of their previous attempt and there are proposals to abolish the 5c coin as well.

The reality is tiny currencies have an economic and social toll far beyond their cost to produce a coin. And there haven't been any negative effects in any countries which have done this.

America is the best in the world right? Why do you think you can't do this basic thing others can?

Comment Re:Trump's penny fixation is stupid (I know, shock (Score 1) 68

You call it stupid without any counterpoint or argument as to why. Common let's hear your reasoning. It's not about cost, it's about social impact as well. Tiny currencies are fucking useless, and exact change isn't needed.

Counterpoint: (and seriously fuck you for making me take Trump's side in anything. Like holy shit I hope you have the worst Sunday imaginable) many first world economies have abandoned tiny fractions of their currencies for good reason and to good effect. It's not just the cost of manufacture, its the total lifecycle cost. Pennies and cents need to be handled, they need to be counted, they need to be transported. Regardless of their value, they need to be minted, and for what? The small value causes people to not carry them and the vast majority of small denominations end up lost or wasted as it doesn't make much sense to hang on to it. One estimate says that some 8% of pennies are simply lost to people throwing those worthless things away.

Even the braindead moron in chief is right about this one thing, and in this case he's only copying policy that other sensible western democracies have already enacted.

Comment Re:Europe already failed with that kind of thinkin (Score 1) 16

Europe didn't fail this thinking, they never did it. There was zero parts standardisation between European car brands outside their direct internal corporate line (i.e. Audi, VW, and SEAT may share parts, but they are one company, at no point did they attempt to standardise the supply chain.)

Actually we can look to Europe for this and find success stories. Several industries have done just this. The ICCA standardised supply chain and industrial equipment across the chemical industry. The IOGP standardised supply chain (mostly driven by Bayer) and industrial equipment across the oil and gas industry (mostly driven by Shell). Both have been a roaring success, driving down costs and simplifying product offerings from suppliers.

Comment Re:Ah yes standardization (Score 1) 16

The article ignores that this will have a knockon effect to industries that support car manufacturers.

Multiple industry have pushed for this in the past and universally the support and supply chain in those industries have been hugely in favour of these kinds of standardisation. It doesn't just make it easier for car companies to specify parts, it makes it far easier for the supply chain to manage inventory of said parts when every company doesn't have some custom feature.

Comment Re:Bluetooth (Score 1) 16

I say the exact opposite. While I agree Android Auto and Car Play are minimum requirements, I will never buy a car whose infotainment system I am not happy with out of the box.

a) Much of the car can't be controlled through AA or CarPlay so the interface needs to be functional for that.
b) Infotainment system integration with instrument cluster is far superior to the limited capabilities AA or Carplay offer. and
c) I never want to be in a situation where I am dependent on a portable easy to loose easy to break device that may or may not have a charged battery or a flaky USB cable.

If the car isn't nice to use, including navigation and UI, then it doesn't get bought. Quite simple. I get it that other people are more concerned with exterior looks over functionality, but it's insane to me that you buy a device that expensive with the intent to bypass parts of it you don't like in advance. What message does that send manufacturers? "Don't bother, users will buy our shit anyway."

Comment Re:Google AI (Score 1) 39

what's there to rave about with Gemini?

It skillfully removed my hand from a photograph I took of my friend's cat. My hand had obscured most of the cat's right ear along with parts of items in the background.

And which model did you use? Be specific, which is kind of my point. Google's image editing AIs are any one of 5 different models which are dynamically selected. What you chose now may not be correct later. How it edits may not be perfectly consistent either.

Nano Banana Pro (Google's flagship image model which probably fixed your hand) is absolutely amazing at photo realistic images for single edits. But it absolutely sucks when it comes to temporal consistency varying an insane number of different parameters. So while Google's AI is great for one off edits, you may find yourself requesting that edit multiple times until you're happy with the results. On the other hand if you want to make an edit to say a photo storyboard Google's AI sucks for it since it can't apply consistent parameters between images. Midjourney is far superior in that regard.

I actually use Google's Image tools the most because they suit my needs. But while it often is able to produce the best looking outputs it fails on two points:
a) It often ignores or doesn't understand referential instructions. e.g. "Put the person's right leg in a cast" or "put the leg closest to the left side of the image into a cast" or "put the person's camera-left left into a cast" has resulted in a crap shoot of 50/50 left leg and right leg results. Internal referencing works e.g. "put the leg closest to the table into a cast", if you have one table. But good luck if there's a table on the left and right because it'll screw it up.
b) It fails to accurately generate occlusion. For example you want to add a picture of your cat half peaking out from behind a building and you describe your cat, it understands cat head, and building, and then can't cope with not drawing both in an almost complete fashion because it assumes everything you've described should not be occluded. It's an even bigger problem for transparent objects. The solution there is to ask Google to make the image twice once describing the cat and the second time the building in front of it. This is an area where Seedream runs circles around Google.

Also you want to sharpen or enhance a picture? GPT's Sunburst model also runs laps around Nano Banana.

Comment Re:Google AI (Score 1) 39

Again of what? Be specific. The AI models vary hugely in capability in different tasks. It's one of the main reasons that CoPilot allows you to select from multiple different models in the first place. They aren't all equally good for the same task.

Do you think Gemini is better than Claude at coding? At summarising the news? At translating languages? If you're claiming that it's simply the bestest at everything everywhere ever, then I'm going to go on a limb and say you've really not been exposed to many AI models.

Comment Re: Space isn't Cold. (Score 1) 131

Well, in a sense he's right. "Cold" and "heat" are not "real absolute thing"s. They're averages of velocities, and velocities are relative.

Velocities are relative, but when two things relative to each other have no velocity on an atomic level there's zero energy available to be transferred and by definition nothing could possibly be colder. That's what I was pointing out. While the idea of cold and hot may be relative, there certainly is an absolute point we can't get beyond (and fun fact, that is much colder than space).

Comment Re: Space isn't Cold. (Score 1) 131

Well let me ask you this: would you at least concede that Musk knows more about deploying tech to space than you do? Because I'm betting you would not, and I think that illustrates the dunning-kruger-ness of it all.

I wouldn't concede this. What I would concede is that Musk knows more about running a business than I do, and that he will have people in his business that do know more about deploying space tech than I do. But there's no evidence that points to him knowing anything about space travel beyond the macroeconomics of getting funding for space tech. In fact quite a few of SpaceX's early failures were directly attributed to Musk attempting to make some engineering decisions. In many cases actual people who knew what they were doing got rid of the stupid ideas.

Comment Re:Sending random things into space... (Score 1) 131

But Google, SpaceX, and China are all putting serious dollars/yuan into it, and I presume they have crunched the numbers.

The world is full of "serious" projects and number crunching offering nothing further than an opportunity to bilk investors, the government, or both. Not every project needs to be remotely logical from a technical perspective and there's a reason the world is full of people spending real money on actually truly stupid ideas.

Yesh I'm sure they have crunched the numbers. Those numbers however have nothing to do with the cost of compute.

Comment Re:Does space have temperature? Really, not. (Score 1) 131

A piece of charcoal will be one temperature, a piece of aluminum would have a different temperature, a piece of quartz yet another temperature.

One of those will represent the black body radiation, the others will objectively not. There's a reason the GP mentioned charcoal specifically.

And that temperature will be different depending on whether it's in sun or shade

The GP's post only had 4 sentences and you seemingly didn't even have enough attention span for that.

Comment Re:Google AI (Score 2) 39

Quality of what? It doesn't integrate with the world's most popular office applications making it useless for most of the corporate world, it lacks a speciality such as producing code like Anthropic does. It doesn't have the same level of integration as competitors beyond Google's own ecosystem, and even there it competes with competitors (the "AI" button on my Android phone doesn't start Gemini).

Outside of replacing my car's voice control and producing search results I didn't ask for, what's there to rave about with Gemini? (And heck it can't even consistently work in my car, often interpreting something simple such as "how much faster is it to take the toll road" as "please read me the Wikipedia entry on the toll project" at least 50% of the time.

Slashdot Top Deals

UNIX enhancements aren't.

Working...