this post was submitted on 10 May 2025
168 points (80.0% liked)

Memes

10403 readers
988 users here now

Post memes here.

A meme is an idea, behavior, or style that spreads by means of imitation from person to person within a culture and often carries symbolic meaning representing a particular phenomenon or theme.

An Internet meme or meme, is a cultural item that is spread via the Internet, often through social media platforms. The name is by the concept of memes proposed by Richard Dawkins in 1972. Internet memes can take various forms, such as images, videos, GIFs, and various other viral sensations.


Laittakaa meemejä tänne.

founded 2 years ago
MODERATORS
all 49 comments
sorted by: hot top controversial new old
[–] hanke@feddit.nu 81 points 3 weeks ago (3 children)
  1. You can't have unbiased AI without unbiased training data.
  2. You can't have unbiased training data without unbiased humans.
  3. unbiased humans don't exist.
[–] Zacryon@feddit.org 13 points 3 weeks ago

If you want AI agents that benefit humanity, you need biased training data and or a bias inducing training process. E.g. an objective like "Improve humanity in an ethical manner" (don't pin me down on that, just a simple example).

For example, even choosing a real environment over a tailored simulated one is already a bias in training data, even though you want to deploy the AI agent in a real setting. That's what you want. Bias can be beneficial. Also if we think about ethical reasoning. An AI agent won't know what ethics are and which are commonly preferred, if you don't introduce such a bias.

[–] Delphia@lemmy.world 5 points 3 weeks ago

The best use case I can think of for "A.I" is an absolute PRIVACY NIGHTMARE (so set that aside for a moment) but I think its the absolute best example.

Traffic and traffic lights. If every set of lights had cameras to track licence plates, cross reference home addresses and travel times for regular trips for literally every vehicle on the road. Variable speed limit signs on major roads and an unbiased "A.I" whose one goal is to make everyones regular trips take as short an amount of time as possible by controlling everything.

If you can make 1,000,000 cars make their trips 5% more efficiently thats like 50,000 cars worth of emisions. Not to mention real world time savings for people.

[–] jerkface@lemmy.ca 4 points 3 weeks ago* (last edited 3 weeks ago) (1 children)

show your work. 1 especially seems suspect. Especially since many AIs are not trained on content like you are imagining, but rather trains itself through experimentation and adversarial networks.

[–] RememberTheApollo_@lemmy.world 11 points 3 weeks ago (1 children)

Even how it trains itself can be biased based on what its instructions are.

[–] jerkface@lemmy.ca -2 points 3 weeks ago* (last edited 3 weeks ago) (1 children)

Yes, and? If you write a bad fitness function, you get an AI that doesn't do what you want. You're just saying, human-written software can have bugs.

[–] xthexder@l.sw0.com 3 points 3 weeks ago

You’re just saying, human-written software can have bugs.

That's pretty much exactly the point they're making. Humans create the training data. Humans aren't perfect, and therefore the AI training data cannot be perfect. The AI will always make mistakes and have biases as long as it's being trained on human data.

[–] Skullgrid@lemmy.world 24 points 3 weeks ago (1 children)

Every sci fi work : oh no, the technology is bad

Reality : the assholes using the tool are making it do bad things

[–] eluvinar@szmer.info 1 points 3 weeks ago (1 children)

There's always assholes and they are always making it do bad things, so the distinction isn't even there. If you don't plan for assholes using the tool to try and do bad things, you're making bad technology

[–] Skullgrid@lemmy.world 1 points 3 weeks ago

you're right, ban fire and knives because they allow arson and murder.

[–] ArchmageAzor@lemmy.world 15 points 3 weeks ago (1 children)

I do think that the best government would be one run by AI.

I do not think the AIs we currently have could run a government, though.

[–] twice_hatch@midwest.social 2 points 3 weeks ago

It wouldn't have the mandate of the people. It wouldn't last very long. I think sortition or parliament could work. Long as it's democratic. It's still a huge leap from how the US does things

[–] amzd@lemmy.world 10 points 3 weeks ago

It’s weird to hold the belief that AI won’t oppress us while showing it that it’s fine to oppress animals as long as you’re smarter

[–] MelastSB@sh.itjust.works 8 points 3 weeks ago (1 children)

Do you have time to talk about our Lord and Saviour Samaritan?

[–] hydrashok@sh.itjust.works 3 points 3 weeks ago

Loved that show.

[–] ThePyroPython@lemmy.world 8 points 3 weeks ago

Literally the plot of every sci-fi show with an "overseer".

[–] Fifrok@discuss.tchncs.de 6 points 3 weeks ago

You really want to give any power over yourself and others to the bias amplification machine?

[–] FuckFascism@lemmy.world 5 points 3 weeks ago

Or super intelligent yogurt

[–] HailSeitan@lemmy.world 4 points 3 weeks ago

My brother in Christ, capitalist markets and the corporations that run things already satisfy the definition of superintelligence

[–] OpenStars@piefed.social 4 points 3 weeks ago

That's just what they want us to think! /s 😜

Wait a minute... oh no no no no no no, that is what they want ~~to sell us~~ us to think! (as they game the system and control the AI, no /s no cap!)

img

[–] Fuzzy_Dunlop@lemm.ee 4 points 3 weeks ago (2 children)

Absolutely.

Every time I hear someone question the safety of self-driving cars, I know they've never been to Philadelphia or NJ.

[–] Natanox@discuss.tchncs.de 13 points 3 weeks ago (2 children)

I mean, the US really isn't a good example for road safety. Even Germany got better drivers, and we like to drive 140-200 kmh. It's a matter of good education, standards and regulations (as always).

In the end self-driving public transport is the way the future of mobility should primarily be imho. Self-driving cars… as long as there always is a steering wheel in case of unexpected circumstances or to move around backyards and stuff it'll probably me fine. Just don't throw technical solutions at cultural problems and expect them to be fixed.

[–] Zacryon@feddit.org 6 points 3 weeks ago

I mean, the US really isn't a good example for road safety. Even Germany got better drivers, and we like to drive 140-200 kmh. It's a matter of good education, standards and regulations (as always).

I didn't want to believe it as well, but it seems to be factually correct, as per this wonderful Wikipedia list.

[–] BuboScandiacus@mander.xyz 4 points 3 weeks ago (1 children)

Ah yes the good regulations on german public roads

[–] boonhet@lemm.ee 5 points 3 weeks ago (1 children)

They're so well regulated that they can safely drive on roads with no speed limit, whereas the US for example has pretty low limits and multiple times the fatal crashes (proportionally to population)

[–] Natanox@discuss.tchncs.de 1 points 3 weeks ago (1 children)

This. Of course it would be even better with limits on the Autobahn, and in fact a majority of people are in favour of such a change (especially if the limit is at 130-140). Our governments are in the pocket of the car industry though, politicians act as if our whole freedom is endangered talking about it (now where do we know that from? 🙃). Things can always be better, but A.I. definitely doesn't improve an absolutely shitty mobility system like the US has (which is basically nothing but cars). If anything it will make shit even more… off the rails. 😏

[–] boonhet@lemm.ee 1 points 3 weeks ago (2 children)

Yeah but then what's the point of visiting Germany as a tourist slash petrolhead?

Jokes aside, I'm of the opinion that existing freedoms are generally best left alone. Besides, Germany has a lower rate than Estonia and we have much lower speed limits. 120 on newly built separated highways in the summer (actually these might have 120 with good conditions in winter too - they have digital signage), 110 on old separated highways and in October or so, they go and collect all the 110 signs and replace them with 100... And up to 90 everywhere else.

There's a good chance the limitless autobahn is actually part of what makes German numbers so good. It just requires stricter training and policing, stricter TÜV and for people to always check their mirrors before switching lanes. And just good lane discipline in general. You don't get that in a lot of Europe. People switch lanes whenever because they're going 10 over the speed limit and can't possibly imagine someone else is going faster than them, potentially very close behind, in the other lane.

PS: traffic fun fact: Did you know that in Latvia, a two lane undivided highway has up to four active lanes? There's the law abiding citizen lanes (known as shoulders in the west) and the BMW/Audi lanes in the middle, marked by the white lines.

[–] Natanox@discuss.tchncs.de 1 points 3 weeks ago* (last edited 3 weeks ago) (1 children)

and for people to always check their mirrors before switching lanes.

Oh, I wish. I don't think your expectation of adapted behaviour is correct on a societal level, and given how many deaths could've prevented by a speed limit… people drastically overestimate their abilities and underestimate the speed and force of impact all the time. If the road is going slow right now or someone missed their exit people will still drive like maniacs. Not to mention that there's also other good reasons for a speed limit, environmental and economical (with ICE cars you don't immediately feel how much more you're paying in money and convenience/time, but EVs will tell you that immediately = more CO², more costs individually and for society, less sane car purchases).

I don't think strict TÜV, training etc. is connected to a lack of speed limit either. It's more of a cultural thing in society, and of course to politics and how well people are off.

I get your opinion about preserving existing freedoms. It's always a balance, however in this case I think this personal freedom to go fast is in no relation to other people's right to save travel, and future generations' right of well-being.

[–] boonhet@lemm.ee 1 points 3 weeks ago (1 children)

The emissions part I'll have to agree on, but safety? Germany is literally among the safest nations to drive in. There's not much lower you can go.

As ICE vehicles get phased out, people will naturally start driving fast less often. EVs force you to stop for much longer when you run out of charge. Driving 2x as fast means making 4x as many stops and the stops aren't 3 minutes with an EV.

[–] Natanox@discuss.tchncs.de 1 points 3 weeks ago (1 children)

The emissions part I’ll have to agree on, but safety? Germany is literally among the safest nations to drive in. There’s not much lower you can go.

Always room for improvement, especially if there's such an unnecessary reason it's "only" 16th place. Regarding EVs: You didn't really acknowledged the economics arguments but simply assume that ICEs will be faced out. Which might still be true even in the worst of all economical situations, however my argument also entailed that people want to buy EVs. Giving ICEs and EVs equal footing in a "free" market is a good thing unless you want offset not doing so with a heavy hand with more regulations. One way or another we need this change asap, so a regulation that's literally everywhere but here, has support from a majority of people and benefits every cause there is makes the most sense if you ask me.

"Top 10 in road safety" also has a nice ring to it.

[–] boonhet@lemm.ee 1 points 3 weeks ago (1 children)

EVs have to be legislated into being the only choice anyway. There's no way around it, they're unfortunately inferior for a lot of people's use cases still. We've grown accustomed to the energy density of fossil fuels and being able to keep cars running out of warranty. A quick look at the replacement battery cost of an original Audi E-Tron will reveal that at this point, EVs are expensive paper weights once out of warranty.

[–] Natanox@discuss.tchncs.de 1 points 3 weeks ago (1 children)

Those problems are being fixed with newer vehicle designs though. The charging speed (the 800V systems seem to reach sub 20 minutes consistently by now), range (450+ km appear to become normal, the more fancy ones even got over 650km) and economics (battery packs become repairable and way more affordable even for a full replacement, especially given the savings in other expenses compared to ICE) seem to check out. The main problems are infrastructure, lack of affordable second-hand options… and honestly that stupid idea that cars should be our primary mode of transportation. Damn, we even got viable alternatives to lithium-ion batteries by now, that technology didn't properly evolve for half a century or sth.

The data on battery longevity is also promising, they stay at usable capacities way longer than expected. Those first generation EVs just age really poorly given they're, well… first generation. The technology is developing at breakneck speeds for the last decade.

On a sidenote, don't let anyone tell you european EVs are somehow worse than chinese (they are not). They're just more expensive due to a fortunate lack of slavery, and generally higher standards of everything in the production chain compared to China.

[–] boonhet@lemm.ee 1 points 3 weeks ago* (last edited 3 weeks ago)

On a sidenote, don’t let anyone tell you european EVs are somehow worse than chinese (they are not). They’re just more expensive due to a fortunate lack of slavery, and generally higher standards of everything in the production chain compared to China.

You forget the lovely government subsidies of key industries in China, they subsidize goods for EXPORT. It's why BYD can sell such high quality cars at such low prices.

I'm still rooting for European EVs, but ffs, Mercedes has completely ruined their exterior designs (interior is subjective - personally I don't like so much screen real estate in a car interior, but other than that they still look nice inside), same for BMW. Audi has apparently somehow stayed just as unreliable with their EVs as their ICEs were. Volvo has the EX90 (bigger than I need and quite expensive) and the EX40 and 30 (both too small), but I did just learn that they're going to start making an ES90, which is more my size. I'd prefer a wagon of course, we'll see if they make an EV90 soon, but for now the ES90 is something to consider.

Edit: As much as I hate the front fascia, the i4 might be the car I should be looking at. I prefer lightly used to brand new and their prices are at a nice level now. Equipment is nice too.

[–] eluvinar@szmer.info 1 points 3 weeks ago (1 children)

There’s a good chance the limitless autobahn is actually part of what makes German numbers so good

There's a chance, but I don't think you argued why would it be a good chance.

It just requires stricter training and policing, stricter TÜV and for people to always check their mirrors before switching lanes.

Changing lanes and overtaking are always some of the most risky moments. It's always going to be much much safer if everybody drives the same speed vs. if you have to dodge because people are going 250 km/h for lulz. If you have the stricter training and policing, you still can improve safety by introducing speed limits.

[–] boonhet@lemm.ee 1 points 3 weeks ago

If you have the stricter training and policing, you still can improve safety by introducing speed limits.

What is going to be the excuse for keeping the stricter training and near authoritarian policing if there are speed limits? Nearly no other country is this anal about who can and can't drive on their roads. Maybe Singapore, since they require you to be a millionaire to even get a car.

[–] 9point6@lemmy.world 4 points 3 weeks ago

I mean TBF, they don't trust the average person in New Jersey to handle a petrol pump—so much so that it's legally prohibited.

I'm not at all surprised that they shouldn't be trusted with the vehicle itself, given that

[–] PacMan@sh.itjust.works 3 points 3 weeks ago

Most CEO jobs and a majority of upper management but those will be the last jobs to be automated

[–] gibmiser@lemmy.world 2 points 3 weeks ago

https://en.m.wikipedia.org/wiki/The_Evitable_Conflict

If nearly perfect computers controlled government.

[–] jaemo@sh.itjust.works 1 points 3 weeks ago

Lol @ everyone imagining that they (baselines) would be able to discern the motives and actions of a superintelligent anything.

[–] ZILtoid1991@lemmy.world 1 points 3 weeks ago

These clowns would be the ones setting up skynet...

[–] BaroqueInMind@lemmy.one -1 points 3 weeks ago
[–] DavidGarcia@feddit.nl -3 points 3 weeks ago (4 children)

AI judges make a lot of sense, that way everyone is treated equally, because eveey judge thinks literally the same way. No corrupt judges, no change in political bias between judges, no more lenient or strict judges that arbitrary decide your fate. How you decide what AI model is your judge is a whole new can of worms, but it definitely has lots of upsides.

[–] qaz@lemmy.world 13 points 3 weeks ago (1 children)

Perhaps when we have real AGI, but I wouldn't want an LLM to decide someone's fate.

[–] nickwitha_k 4 points 3 weeks ago (1 children)

You have been found guilty of jaywalking. I hearby sentence you to 90 days of community service as unicorn titty sprinkles from Valhalla. May Chester have mercy on your handkerchief.

  • JudgeGPT, probably
[–] qaz@lemmy.world 4 points 3 weeks ago* (last edited 3 weeks ago)

You could get away with murder if your lawyer talked the charges out of it's context token limit.

[–] biggerbogboy@sh.itjust.works 2 points 3 weeks ago

And how will this be done? A proper legal system needs impartiality, which an AI still varies as much or more than a human judge. Not to mention, the way it's trained, the training data itself, if there are updates to it or not, how much it thinks, how it orders juries and parties, etc.

If, in theory, we have a perfect AI judge model, how should it be hosted? Self host it? Would be pretty expensive if it needs to be able to keep up. It would have to be re-trained to recognise new legislation or understand removals or amendments of laws. The security of it? If it needs to be swapped out often, it would need internet access to update itself, but that produces risk for cyber attacks, so maybe done through an intranet instead?

This requires a lot of funding, infrastructural changes and tons of maintenance in the best case scenario where the model is perfect and already developed. There would be millions, or ideally, billions in funding to produce anything remotely of quality.

All I see are downsides.

[–] untakenusername@sh.itjust.works 2 points 3 weeks ago* (last edited 3 weeks ago)

if it's designed right, it would be great. otherwise it would suck

[–] Fifrok@discuss.tchncs.de 0 points 3 weeks ago

I mean this with the biggest offence possible: AI judges make no sense, atleast with the current way of doing AI (LLMs). It's been known for years that they amplify any bias in their training data. You are black? Higher chance of going to prison and longer serving time. Getting divorced and are male? Your ass is NOT getting custody. Hell, even without that the LLM might just hallucinate some crime not in the data for a case and give you a life time prison sentence. And if you somehow manage to avoid all that, what's stopping somebody from just shadowprompting it and getting the judgement they want? It would also be an easy target for corruption, the goverment wants their poltical rivals gone? Tweak the model so it's just that bit harsher, or just a bit more alligned with some other interpretatnion of the law.

Who would even choose the training data? The judges? Why would they, it would be better for them to sabotage and keep their jobs. Some goverment agency then? Don't want to do that, or you're gonna find out separation of power has a reason.

Bad idea.