Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[NJW2000]

Never really understood AI risk. There's the "weak AI", which is targeted and really just like a very strong linear processing system like the world has been working towards for the past century, a tool used by humans. Then there's the "strong AI" necessary for the singularity, which is self-aware, humanlike, capable of working on many levels and  suffers the key ontological flaw of not existing.

I'm a little biased though, as I'm interested in the Effective Altruism movement, and entertain major doubts about the AI side of that.

Still more worried about the environment and nuclear warheads, issues that demonstrably could lead to the apocalypse.

[Tack]

Oh yeah I'm mostly worried about the humans behind the AI.
I guess so is Google or Musk or whoever is trying to develop that freeware AI.

[ChairmanPoo]

Spoiler (click to show/hide)

[Akura]

↑...That's pretty soul-crushing right there.


Over a foot of snow, and it's still coming down. Not sure how plowed my street is. I'm starting to hope work calls and says not to come in.

[Ultimuh]

Over a foot of snow, and it's still coming down. Not sure how plowed my street is. I'm starting to hope work calls and says not to come in.

*laughs in Swedish*
*cries in Swedish*

[Truean]

I've explained the numerous points behind that simple, harmless request in detail so many times that other people know it very well.

[Loud Whispers]

Tell me about it. I had one of my threads derailed by people trying to convince me away from obliterating all AI at first sight simply because I RP'd a cautious approach to AI, even in context they could not separate my RP from their assumptions upon my personal beliefs

Also a self-driving car killed someone recently so it's topical

[Putnam]

Consent? Individual preferences? No, those things don't matter. Empathy is uncool and the idea that conversation on the internet involves other people has mostly gone by the wayside in favor of the idea that you're yelling at an audience. Asking something of someone else is infringing on their right to never, ever change their behavior even in the slightest.

[TamerVirus]

Other people? Nah, y’all just figments of my imagination /solipsism

[Loud Whispers]

UK sending shitposter to prison for shitposting

The judge ruled that he wasn't funny enough to not be imprisoned, because the Scottish Judiciary is the best judge of comedy. Jk lol the UK is fucked, no liberty

[Tack]

So what was the original video? Did he just shout “sieg heil!” at his dog and it would raise a paw?
Oh dear how terrible

[Maximum Spin]

UK sending shitposter to prison for shitposting

The judge ruled that he wasn't funny enough to not be imprisoned, because the Scottish Judiciary is the best judge of comedy. Jk lol the UK is fucked, no liberty

this sort of thing is why I say the UK is a third-world country :|
(probably better to let the AI conversation die, but this was so petulant that I can't ignore it...)

I can talk to a human being; can the AI be persuaded, pleaded with? 
If an AI is wrong, is there any recourse? Pedestrians killed by humans get lawsuits; a self driving car kills and nobody's at fault?
Others have picked up on the "Just a tool," "neither good nor bad inherently" argument. Sadly, many think machines are infallible.

Persuaded, yes, with good evidence. Pleaded with, no, but that's the point. Pleading is dangerous. Emotion is dangerous and should be eliminated to every extent possible. Remember, a human can be persuaded or pleaded with to kill you just as easily as not to.
Further, most of the time, when pedestrians are killed by humans, nobody is at fault either, and any lawsuit would be purely frivolous. Well, to be frank, I would wager that most of the time the pedestrian is at fault, but the pedestrian's recklessness has already been repaid justly. If a self-driving car does not behave recklessly, nobody is at fault; if it does, the programmers are. Still, lawsuits are a terrible form of "recourse" and it would be preferable to obviate them altogether.
To sum up, while machines are certainly not infallible, they are at least predictable, which is a fair sight better than humans are.

[Putnam]

you were told twice in the previous page not to quote truean, man, come on

[IcyTea31]

UK sending shitposter to prison for shitposting

The judge ruled that he wasn't funny enough to not be imprisoned, because the Scottish Judiciary is the best judge of comedy. Jk lol the UK is fucked, no liberty

Story from 2011, Finnish. In 1941, the Nazis were infuriated by a Finnish businessman's and his German wife's (known to have been against the Nazis) dog that did the nazi salute whenever someone said 'Hitler'. A lawsuit was threatened, but fell apart to a lack of willing witnesses. They also threatened to drive his business to bankrupcy by refusing to export, but that didn't go through either. Today the historian writing the report wonders why the Nazis even cared enough to spend precious time on this while they were preparing to invade the Soviets.

[Maximum Spin]

you were told twice in the previous page not to quote truean, man, come on

Oh, is that the same guy?

I didn't catch up on the thread between when I last posted and now so I have no idea what you're talking about, but Tiruin did say something about somebody not being quoted in Discord and I'm like "uh, I don't know that guy from Adam".

I'm not comfortable with changing my post, but I'll try to keep that in mind.