Will AI Destroy Humanity? A Soho Forum Debate

[@CurtHowland]

I would like to propose a new Logical Fallacy: "Appeal To Friedman"

[@jacobnussbaum2309]

Even Friedman said that wasn't a fallacy

[@СергейМакеев-ж2н]

I'm really disappointed that the guy who was arguing "against" doesn't believe in any AI at all. We really need an in-depth discussion about how to solve the problem, _assuming the problem exists._ Especially about the question of, "Would state regulation do any better than private means?"

[@avigator]

Actually, he is earning very good money with his AI company. 🤑

[@VeniVidiVid]

When have politicians and bureaucrats EVER been more responsive, nuanced and adaptive than millions or billions of market participants? If AI is used to commit crimes currently on the books, prosecute the harm from those crimes, rather than prior restraint on the tool itself. And the Luddite claim that lower skilled jobs will disappear and cause mass unemployment has been repeated for hundreds of years, and is always wrong. Jobs change all the time. There are fewer buggy whip manufacturers or elevator operators these days. Did those people simply die in poverty? Or do people just have different jobs today?

[@donaldhobson8873]

If AI goes straight from "nothing obviously wrong" to "everyone drops dead", you don't have much chance to prosecute crimes. In general, any scenario where the AI gets out of hand quickly is one where the "prosecute those crimes" approach fails.

[@VeniVidiVid]

@@donaldhobson8873 Same can be said for power tools. Or fire. Humans have used potentially dangerous tools since we’ve been humans. And they do occasionally take their toll. In those cases, we learn how to use them more safely. Apocalyptic claims for AI should probably be weighted with more evidence and specifics than just speculation of unspecified outcomes. But I’d agree, it would be extremely foolish to put AI in unsupervised control of weapons like missile systems.

[@СергейМакеев-ж2н]

@@VeniVidiVid If it was possible to create a kind of fire *that never ever stops burning on its own* (before we invented the appropriate extinguisher), then it would have been analogous. It seems to me that the apocalyptic claims are based on the idea that, once an AI has a "wrong" goal, it will never stop trying to accomplish that goal. And we will have no idea how to stop it either. Doesn't mean that the "fire extinguisher" can't exist in principle, we have just not invented it yet. But then, there's a counterargument that maybe strong AI is _exactly what we need_ to develop countermeasures against strong AI. That thesis totally might be true, but it needs further evidence.

[@onceamoth]

Your assumption that there is a law of nature that gives humans a right to eternal employment might prove to be stuck in the past one day.

[@DanielVoglerTime]

Really wanted to listen to this but simply couldn’t. Her voice had me running 🏃

[@El_Hicks]

She seems to not be conscious that she's screaming for no reason

[@jdmchoes92]

this is what has me disconnecting from her message as every time she speaks it sounds like she is a news reporter trying to overdramaticize everything.

[@ibbajibbaduay]

After-party must have been wild

[@Mar00f]

😂😂 you know it

[@dadsonworldwide3238]

Reminds me of the royal institute scientist claiming flat earth to sway sailors from exploring the new world while they was racing to colonize and control it lol

[@mlcorcor]

This might be the first debate in Soho Forum history where both presenters argued in support of the resolution, yet the no vote won out in the end.

[@iCanSeeWhatMostCant]

AI is going tonkill us all!!! The solution is more government!! 😂

[@Gr88tful]

1:16:59 100% 👏

[@Redact63Lluks]

AI is going to make so many data processing jobs obsolete. AI ain't going to run weld, frame homes, or plumb electrical. I'm all for it removing desk jobs, which it will do fantastically.

[@Gr88tful]

Not immediately. Tbh I do see in the "own nothing be happy" with planned cities future - the power, the water..... Everything - control by a panel/board- o an ai plugs into it and addresses and fixes the problem as well as controlling the use ...

[@avigator]

Susan Schneider’s high pitched voice is a clear indication, that she actually doesn’t know what she’s talking about.

[@Corteum]

_"Will AI Destroy Humanity?"_ Sort of like asking "Will the printing press destroy humanity?", or "Will the industrial revolution destrroy humanity??", "Will the intenet destrroy humanity???" ----- Nope. None of those. But asking stup|d questions might ! Lol

[@Smokedouttasian]

I think it's more comparable to nukes but I see your point. That being said we should be proactive in implementing laws in reducing the damage AI can do. Such as Hacking techniques with AI software which is probably the biggest problem we have

[@Corteum]

@@Smokedouttasian Yes. Hacking is one example. Also social engineering; things like manipulating discourse on forums, comments sections, soc media platforms, manipulating ppls news feeds, etc. That said, I'm not sure how we're going to be proactive in implementing laws to mitigate against potential damage to individuals and societey when we have a bunch of id|otic politicians... w/ shxt for brains i might add... trying to decide what's best for us all. I mean look at what this current administration is tryring to push? Search for the clip titled: "Whitehouse proposes IDIOTIC AI Regulation!" and have a watch. 👍

[@donaldhobson8873]

So a world where superhuman AI designs a bionano weapon that kills all humans, that just can't happen according to you? If there was an AI that was substantially better at inventing new weapons (and various other things) than humans, and it wanted to kill all humans, you confidently predict it wouldn't manage?

[@zwerko]

@@donaldhobson8873 We don't need AI to destroy ourselves-we're more than capable of doing it right now, all it takes is a few insane individuals at the right positions of power. I'd argue it's more likely to happen than (G)AI coming up with a motivation and a plan to 'destroy all humans'.

[@Corteum]

@@donaldhobson8873 A poorly designed one might do that, yes. But how do you propose that we stop it? By letting idiot|c politicians decide what to do? That's not an option.

[@apemancommeth8087]

I throughly enjoyed this detailed, educational debate about A.I! Science & Technology debates that explore where we are in detail is so much fun, but not as fun as where we might be headed 🤔😀🚀

[@pfarabee]

😂 This man's argument basically amounted to "It is ludicrous to consider the negative implications of flying machines, because man cannot fly and will never be able to."

[@alexanderx33]

Wise Old Bird: Our world suffered two blights. One was the blight of the robots. Arthur: Tried to take over did they? Wise Old Bird: Oh my dear fellow, no, no,nonono. Much worse than that. They told us they liked us. Of course not their fault poor things, they'd been programmed too, but you can imagine how we felt, or at least our ancestors. Arthur: gastly. The Wise old Bird: Precisley!

[@drdaverob]

Is this a hhgttg reference? I'll thumbs up for that!

[@DarkHorseSki]

Friedman did not support UBI, but something that did give back to those that work (a negative income tax per se.)

[@alexanderx33]

Jesus she probably practiced talking loud so all the geriatrics in congress could actually hear her.

[@wolflarson71]

Great debate. Smart panelists.

[@jatelitherius9842]

I wish Schneider would have actually argued for the dangers we face. As for the fellow arguing against, even more so I wish he would hang this claim that we can never build intelligence. Maybe we can’t replicate human intelligence- but we can make something which approximates it so well, or is different in key ways, & is still dangerous. Novel solutions & fore planning indeed, are already upon us. Consciousness, self awareness, these are not necessary for danger

[@magicmon]

I can see AI causing alot trouble for someone whom crosses the Gov.

[@VeniVidiVid]

1:27:40 The US federal government has to write new laws…to keep up with the Chinese??

[@dariusthurman8835]

Regarding Milton Friedman on UBI; He wanted all welfare programs rolled into one Super Welfare not an entire Justice League of WELFARE.

[@stevecooper3010]

Nah mate We're more than capable ourselves Take a look at the world leaders and where we are at

[@homewall744]

There is no reason why intelligence must be modeled on a brain of anything. All it has to do is accept input and make useful predictions about the future.

[@Rob337_aka_CancelProof]

It's unfortunate but every time I think about this the song "right in to" from Tool comes to mind because there's always somebody out there who will do something just because they can even without malicious intent if something is possible despite how improbable it may be you can count on it happening eventually and it's hugely irresponsible to just put these potentially dangerous increase the more complex models out there in the public space for anyone to have access to and that's been done since day one. Every single person who's ever done a captcha has contributed to this problem everyone's complicit and I'm not saying everybody's malicious I'm just saying this will result in many many many many many many unintentional self-inflicted wounds that just aren't worth all the potential inevitable problems. This race to the bottom we've been on lately does it need its efficiency elevated buy an order of magnitude we're doing bad enough all by ourselves and we just simply don't need any help where that's concerned.

[@homewall744]

Why can't AI create new weapons, new viruses, new diseases, new ways to commit international problems that cause humans to over-react.

[@4363HASHMI]

Yes

[@mustang607]

Inhumanity can destroy humanity, although AI can be an effective any means tool to do so.

[@homewall744]

Be more afraid of human minds. Some are evil. Some want to control others. Some are fine with violence, lies, theft and extortion.

[@tallard666]

What a silly question. Humanity has already been destroyed, by civilisation and tech.

[@science212]

AI is a good subject. Evil computer is just idiot cyberpunk fiction. Conscious computer is possible, but just theory.

[@donaldhobson8873]

The problem isn't so much that it's evil, it's the AI doing exactly what it was programmed to, but in a way we don't want. A combination of software bug and be careful what you wish for.

[@iCanSeeWhatMostCant]

@@donaldhobson8873 for example, the AI is programmed to prevent the extinction of the human race, it completes that by killing 95% of humans and then imprisoning the last 5% underground for eternity.

[@chemicalwasted3450]

It sounds like she wants funding for her institute from Congress.

[@avigator]

Susan Schneider’s high pitched voice is a clear indication, that she actually doesn’t know what she’s talking about.

[@standingbear998]

it will do whatever it is created and programmed to do. When it is acceptable for people to openly say they want millions to die for depopulation? don't be distracted and focus on the tools instead of the murderers

[@brianmahoney4156]

this guy is so off base. “all we have is a syntactic algorithm that can reproduce what its seen before.” no, you dont know that. you dont know what the algorithm is. its 100 percent a false claim. its much more speculative than the claim that these systems might become dangerous.

[@tann_man]

Why is this guy so stuck up on the definition of intelligence. That's hardly important. The question should be do these algorithms pose a extisential risk and if they do ought they be regulated and if they ought how exactly should they be regulated.

[@donaldhobson8873]

It will destroy humanity if we don't manage to stop it doing so. Ie if we don't bother with AI safety, we are doomed. Will the AI safety work? I don't know.

[@GiRR007]

No it wont. It has no reason to. The very idea of AI going out of its way to destroy humanity is anthropomorphism. Its just a tool like any other.

[@СергейМакеев-ж2н]

You remind me of a phrase that Robert Miles once used, it was something like: "At least this design, if we manage to implement it correctly, does _slightly better_ than just "everyone is immediately dead"".

[@davidrussell8203]

Elon Musk has built a robot for ma's distribution.

[@jarrodyuki7081]

ai geen drives nad thoriu mreactors.

[@Invisibility397]

Hell yeah AI is Far more Logical than Women that Ghosts and Abandoned Young Men. So yes women will definitely never understand what Character Values Morals Standards and Principles are. Respect Loyalty Honesty Dignity Integrity? Right. Stay making life decisions based on Temporary Emotions

[@patricksullivan4329]

I'd say this is one of the worst cases of 'the curse of knowledge' I've ever encountered, except it's two. Neither of these people have a clue as to how to present ideas. (Hint: don't throw jargon around and expect anyone to understand what you are saying). One thing I'll recommend to the scatter-brained blonde; drop the gratuitous reference to Milton Friedman and read some of his U of Chicago colleague George Stigler's work on 'regulatory capture'. I.e., Quis custodiet, ipsos custodes?

[@avigator]

Susan Schneider’s high pitched voice is a clear indication, that she actually doesn’t know what she’s talking about.

[@drdaverob]

Omg I can't stand her voice. Misophonia, ppl. It's a problem for some people. Look into it.

[@riclacy3796]

The number of folks who are confident about the lack of threat seems quite alarming by itself. New tech can be dangerous and disruptive - let's just hope it's only as bad as the printing press and not as bad as nuclear energy.

[@donaldhobson8873]

As someone who knows a lot about AI, I would be astonished if it was only as bad as nuclear energy. I mean we absolutely could get nothing seriously bad happen by luck and wisdom. But the inherent technology danger of superhumanly smart AI being lower than the inherent danger of nuclear energy, no way is that true. It would be highly surprising if superhuman AI turned out to be less dangerous than nuclear weapons.

[@Monkey-fv2km]

But nuclear energy I assume you mean nuclear weapons, since statistically speaking nuclear energy is as safe as solar energy.

[@stoya2s]

Of all the varieties of the English language, the northern American accent has to be the most unpleasant to the ear. What the fuck is "darn"? Why can't you just say "damn"? What if we randomly start changing all the words we dislike with nonsensical alternatives?

[@drdaverob]

North American accent? Bahahaha. She is not a good example for any region in America. And you're not talking about accents you're talking about dialect or vernacular of interjections. If you just dislike Americans say so, but don't say wrong things to justify yourself.