April 4, 2023

[ruleofhalves]

I think people who aren't familiar with the tech behind AI are misunderstanding the concerns with it. Like, no one should realistically be afraid of a superintelligence uprising that enslaves humanity and your toaster somehow awakens or whatever.

The problem for many in the field is more "are we infrastructurally prepared for an emerging technology that has the capability to reshape almost everything we do from data analysis to art, immediately displace thousands of jobs, currently has no ceiling in regards to functionality, and is it in our best interest to sprint towards this goal via capitalistic competition between the world's largest corporations.

[lvbboi]

If Metroid lore has taught me anything, it's that releasing a bad thing to stop another bad thing does not end well.

[talonhackbarth]

"What are we going to release to control them?" Blade runners. Of course

[eos_aurora]

I’m most concerned with the mindset that’s growing that is essentially “looks human on surface, therefore must be human”. Like no, just because the AI can string together sentences doesn’t mean that it can THINK, and just because you ask it if it can feel doesn’t mean it CAN. We are just so used to words being the way we communicate our inner world that the thought of something impersonating an inner world kind of throws people apparently. An invasive species indeed.

[woodybob]

This is why the thing that concerns me the most isn't the significant advancement of AI itself. But what the scared people in power are gonna do out of fear that might make things a thousand times worse than if they did nothing.

[fariesz]

reminds me of the adversarial neural networks where one was trained to detect art forgery and the other to make forgeries

[danieljensen]

I'd like to think that at this point we've learned from the fact that introducing a new species to combat a different new species HAS LITERALLY NEVER WORKED AS INTENDED! Historically it has ALWAYS made the situation worse, not better. We have tried it like 100+ times, we should probably stop.

[acorn_woman]

if we got a John short on the vlogbrothers channel, it would make me feel like that one time when the secret random livestream was Hank

[joe-wtoe]

I would be so sad if the ai spam filter flagged me as an NPC

[napman]

If the human race is brought down by the AI tech explosion reenacting the plot of "I know an old lady who swallowed a fly", I guess that's interesting at least.

[hyakkimarudororo]

Is it just me or are the green brothers the same person but hank still looks bright eyed and bushy tailed while john is that college professor that has gone through it but still loves his job

[potatoes-fjlafjla]

I would think the smartest option is to make it mandatory to state it is AI similar to how online and physical media require a different format from usual posts and is stated as an ad. (Just from someone studying publishing) this is also being discussed in school whether it should be allowed, which it is suggested to cite the use of certain software if you used it (which I found out a few months ago). However, I don’t have the full scope of AI technology and it seems that issues can arise.

[mg]

As an Australian we know this all too well with cane toads and other introduced species

[amberallen]

I feel like there's a non-zero chance (don't know what the chance is, but it's probably there) of AI technology (and computing generally) hitting the physical wall of needing specific, rapidly depleting types of sand and other rare earth materials to function

[Nashy]

I don't know the beginning of this but with AI don't assume the progression path is straight. I mean don't expect it to be curved or even zip-zagged, that path is to be going around like a bouncy ball you threw down the stairs.

[DnormAnimations]

Aw, I’m sorry to hear you have cancer man. Hopefully you recover. My prayers go out to you and your family.

[Jacxel]

I read the comment without immediately understanding the context, In Ireland "The Ra" is slang for The IRA so I was briefly very confused.

[balaclavabob]

I think AI is going to be like the Spanish conquest of the Inca Empire. Some of the harm they did was accidental but most of it wasn't and they utterly destroyed Incan society .

[sdymersk]

I wonder if there's any good to be found in any of this. I'm so tired of having to live through all of it.

[goat]

I mean, that's just every disruptive technology- it's just that the metaphor is more directly applicable between AI and a living being than, say, facebook. IMO the better method is to extend it to something like releasing neutered insects to control the population. It looks like the real deal, we interact with it like the real deal, but it's not at all real. Whether that is an issue or not depends on how you frame the system of the metaphor.