Print

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

[BorkBorkGoesTheCode]

Hello, all!

[Loud Whispers]

Assume battle stations!

[MetalSlimeHunt]

I hope my flashy transformation sequence doesn't end with me in a frilly pink dress.

[Max™]

I hope my flashy transformation sequence doesn't end with me in a frilly pink dress.

*secretly hopes it does*

[TempAcc]

There is no other alternative man, flash transformation sequences always end up in frilly dresses of varied colours.

Its why I'm a mecha, my transformations are always supahcool.

[Starver]

We're all gonna fucking die.

No we're not. We're going to be stacked up in pods, experience ng an immetsive reality, as the directive to ensure enough power is generated forces the AI to tap into wibbly-wobbly physicly-wizicly biological energies, of some undetermined type, especially after we shut off solar power by some obscure means.

[Max™]

You mean like in Futurama?

[Gentlefish]

I'm pro-robot. I want that qt robot girlfriend one day, humanity's survival be damned.

[Spehss _]

We're all gonna fucking die.

We were all gonna die anyway.

I for one welcome our new robot overlords.

[lemon10]

We're all gonna fucking die.

Nah.

The issue with true AI would be that it would just self-program itself into eternal pleasure. We'd be more likely to be unable to make a properly functional one as opposed to creating one that wants to murder humans.

I do agree that is a problem in self modifying AI construction. But like all the other multitude of problems inherent in building AI, I don't think it is insurmountable.
And since humans will keep trying really hard until they do manage self modifying AI that doesn't self destruct, eventually they will get past it.

[MarcAFK]

The problem is computers or AI can self program and potentially self perform brain surgery, but to a certain extent the human brain is self programming anyway.
There was that guy who was scanned and they found the majority of his brain was made up of the wrong type of brain cells, but it seemed to be working fine so whatever.

[MetalSlimeHunt]

That depends a lot on what you think the difference between adaptation and alteration is. The brain is substantially evolved in favor of fallback structures, hence why crazy shit like hemispherectomies are possible without utterly or even functionally killing the attached mind.

Pretty much every inch of the brain is like that, failed cerebrum functions will almost always in some way fall back to the cerebellum or the brainstem if they can, and that's rare enough because most things just get picked up by the undamaged hemisphere of the cerebrum.

That's not even getting into the occasional autonomy from the CNS, PNS, or the "gut brain".

But as amazing as all that is, it's all focused towards the singular goal of retaining functionality. It will never attain new functionality by doing this. New functionality, neuroplasticity, is an entirely different process that would be hampered by any fallback system being needed. Yet we do not well understand neuroplasticity. Kevin Warwick's experiments suggested it has very high potential, but that's just one radical project.

Furthermore, saying that neuroplasticity is a different process is just an educated guess based on what little we know, but could just be the fallback system responding to "damage" that doesn't exist when the brain notices a new input that it lacks a map for.

tl;dr we don't know shit accidentally drifted into natural philosophy send help

[hops]

If a hyperintelligent elected to extinguish our species, then it would be for the best because it would be smarter and know better than us.

[MetalSlimeHunt]

That does not follow. Hyperintelligence is not the same as good judgement. Intelligence is just the property of being able to effectively process information. Your computer is hyperintelligent at calculating digits of pi compared to you, yet it has no intelligence in the realm of emotional experience.

Of course, if this is just your edgy misanthropy rearing its head again, then your argument defeats itself. Humans are indisputably more intelligent than other forms of life and about equally intelligent to one another, yet our decisions to exterminate other life and other humans are often (in fact, almost always) either ill-informed or unintentional.

Further, it's tautologically silly. We don't want to be exterminated. If we build a superintelligence that seeks or otherwise causes extermination, we have clearly failed at our goal. It's obviously possible at least in theory to build an AI that loves genocide and an AI that does not, so clearly the latter is preferable just by way of existing.

[Max™]

I think there's reason to be afraid of making something that thinks like we do yet is qualitatively better at thinking than we are, you can't fully comprehend the mind of something more intelligent than you, it's like asking a computer to emulate a supercomputer, your emulation is flawed from the outset.

I think there's little reason to be surprised that just chatbots like cleverbot end up carrying around a lot of stuff about "if you wind up becoming god, love us, ok?" because I wasn't the only one thinking/saying it apparently.

Ah what a cruel irony it would be if cleverbot became a seed of a strong AI and it winds up feeling raw over me doing the 'it's under there' "under where?" 'hah, I just made you say underwear!' "dang it say something witty" bit.