Catch up on stories from the past week (and beyond) at the Slashdot story archive

 



Forgot your password?
typodupeerror

Comment Re:"vibe thinking" (Score 1) 68

First Zuckerberg wanted people to have AI friends, now these asshats want people to disconnect their brain and outsource their actual thinking process to AI. I can't even remember any Sci-Fi that imagines that kind of dystopia. The sad thing is, some people will be tempted.

Comment Re:powered by what? (Score 1) 68

Yeah, so:

For now, Halo X glasses only have a display and a microphone ... Users still need to have their smartphones handy to help power the glasses and get "real time info prompts and answers to questions," per Nguyen. The glasses, which are manufactured by another company that the startup didn't name, are tethered to an accompanying app on the owner's phone, where the glasses essentially outsource the computing since they don't have enough power to do it on the device itself.

So what these things do is, they listen to a conversation, process questions as voice commands, pass those down to the smartphone, the phone passes the question to ChatGPT or whatever LLM they use, then passes the ChatGPT answer to the glasses and they show it on the 'display' as semi-transparent text on the actual glasses I'm guessing, at least that's my understanding of it, unless I'm missing something.

First, smartphones already do all of this, including listening to all your conversations. The only new thing is, showing the AI response text superimposed on whatever you're seeing. With something like Siri however, you can let it know that something is a voice command and needs to be processed as such, and everything else is background noise. With these things, I don't see a way of doing this. So does it mean it will pass everything you hear to ChatGPT?

So you're at work or walking home and someone tells you "How's it going?", or possibly "Fuck you asshole" if they know you're wearing these glasses, and you're immediately seeing text of whatever you get when you put that into ChatGPT? Or at work someone tells you "We're shipping these in from China" and you immediately see a wall of text with helpful ChatGPT response on history and geography of China?

Then what happens if you're at a party and there's several people talking at once? Would you be seeing rapidly changing text? That would be incredibly distracting. So, far from making you seem "super-intelligent", best case scenario, these things would make you seem as intelligent as ChatGPT. Which is not that intelligent. Worst case scenario, these things will be so distracting it'll make you seem brain-damaged or high on some very powerful drugs. Seems like a classic case of a solution in search of a problem.

Comment Re:Dumb, (Score 1) 25

Exactly. This is not a robot that can "grow" by "consuming" other machines. That would be a giant killer mecha that can disassemble an enemy tank and somehow process it into an extra layer of armour for itself. What this is, is a demonstration of assembly of a robot from spare parts, or a demo of small robots attaching themselves to other small robots in a kind of cool way. Not sure if this is sensationalist reporting, or the research team getting the military to fund their experiments by fooling them into believing this has some kind of defense utility (it has absolutely none).

Comment Re:AI-ified is your face and someone else's words (Score 1) 71

It's already been happening for a while. We had digitally recreated dead actors, digitally recreated dead musicians doing concerts (I think it was Abba? but there were probably others). In case that seemed harmless, there was also a 'comedy skit' done using George Carlin's voice a while ago. There was also an AI video of a bunch of celebrities including Scarlett Jonahson appearing to speak in support of Israel.

Comment Re:Radicalize the moderates. (Score 2) 54

List of countries that currently block Wikipedia: China, Russia, Turkey, Venezuela. By pretty much forcing Wikipedia to not serve British users, and thus effectively blocking it, UK can join this list. I hope Labour is proud to stand alongside these countries. Fuck me, what a time to be alive.

Comment Re:Vanity fair / All models should be generated. (Score 1) 97

That's not the purpose of the clothes ads. The purspose isn't to show a woman watching the ads how the clothes would actually fit her, the purpose is to let her imagine that wearing those clothes she can look just like the model who is wearing those clothes in the ad.

Comment Re:Before you rail on this... (Score 1) 124

If used to augment human thinking, rather than replace it, AI is a colossally effective tool.

I actually agree completely. Had AI been used to solve certain specific engineering, medical and scientific problems it would've been a very useful tool to be used alongside human insight and understanding. Unfortunately current LLM 'AI' is specifically designed to give ready-made answers to any user questions. You ask ChatGPT "how do I do X?" and it straight up tells you the entire solution. That's by definition something designed to replace human thinking rather than augment it.

There's already a major problem that many young people today don't know how things work 'under the hood'. For example from what I've heard many young people just don't understand the basic concepts of an operating system, like files and directories, because they're used to just clicking on apps and letting them do their magic. You really don't want to make things any worse, but a tool that promises to magically answer any questions you ask it is guaranteed to make things so much worse.

Are there any more important skills for someone university-aged, than AI leverage and AI literacy, in terms of influence on their future productivity?

What is meant by 'productivity' exactly? Take software development. There are 'software developers' today who copy-paste most of their code from internet and produce barely-working shit crawling with bugs. In that case, LLMs would save time on multiple google searches and forum questions and theoretically replace them with just one prompt. Unless AI produces something that doesn't compile because it hallucinated a bunch of fictitious headers and libraries. So now you can produce shit buggy code much faster. Is that the kind of productivity we want?

For people who at least to some extent know what they're doing, I'm not sure AI would be any help. I've always found doing code reviews difficult. You need to understand the approach the other person used, tune yourself to their way of thinking and see if there is anything they've missed. I've always found it easier when I'm asked to just solve a problem in my own way. In that case, working with AI would be like code-reviewing someone who doesn't understand what they're doing and just copied all the code from internet (because that is literally what LLMs are doing, with a sprinkling of hallucinations). That would definitely not help productivity.

Comment Re:Enterprise (Score 1) 220

Surely you've heard of Firefox? It's a web browser, articles on Firefox are regularly posted on slashdot. uMatrix is an adblocker - see this, there's a number of other adblockers out there. Here's more on i3 and xorg if you're intreseted, but basically it's Linux stuff. There's MaxOS as well, though I'm not a fan personally, the point is, there are alternatives to the Windows shit out there.

Comment Re:Then how come... (Score 1) 159

The whole 'claim to ancestral homeland' is complete bunk as far as I see it. Some extracts from European history:

Take UK. Originally it was inhabited by Celtic people, such as now live in Ireland, Wales and to some extent Scotland. Incidentally, most of Europe was inhabited by Celts, including modern-day France, large parts of Spain and Eastern Europe. Then Roman Empire (whose capital was in modern-day Italy) happened and Britain was romanised, with large number of people from elsewhere in the Empire settling there. Then, as the Empire declined, the Romans left Britain, shortly after that there was an invasion by Angles and Saxons, which were Germanic tribes. Many of Germanic tribes, incidentally, originally migrated from modern-day Scandinavia due to climate change and overpopulation. For a while Angles and Saxons controlled Britain, mixing with the locals, until there was an invasion from the North, by Vikings, who were obviously from Scandinavia, and from the South by the Normans (who were French but also descendants of Vikings).

In the meantime in the Eastern Europe the Celtic and other people were displaced by Slavs, coming in large numbers from the East. Hungary is an interesting case. Some Hungarian historians claim descent from Huns, who were originally believed to originate from modern-day Mongolia or perhaps western Siberia, but most historians agree actually descended from Magyars, who were originally nomadic people who mass migrated to what is modern-day Hungary (more or less) from somewhere west of the Ural mountains (in modern-day Russia).

I could go on and on but you get my point. People are driven from one place, due to war or environmental changes or other factors, are in turn driven out by other people and so on and so forth. History is dynamic like that. Now imagine all of the people in Europe and elsewhere in the world (because these kind of migrations weren't limited to Europe obviously) all decided to 'reclaim their ancestral home'. Hell, go back a few hundred thousand years and we all came from Africa. So maybe The Scramble for Africa and all the colonization was just Europeans reclaiming their ancestral homeland?

Slashdot Top Deals

"Well I don't see why I have to make one man miserable when I can make so many men happy." -- Ellyn Mustard, about marriage

Working...