'AI is already able to manipulate people': Yuval Noah Harari warns of growing AI power

[@OctaveIndustries]

All of you saying "education" is going to solve this problem? Sorry to break it to you, AI is going to outpace education exponentially. Learning is what it's best at.

[@Merriwether-w8k]

True - education will not do it

[@dapsolita]

​@@Merriwether-w8k...unless we adjust our notion of what education is.

[@nottiification]

The fact that Americans are so easily manipulated doesnt speak to our need to regulate AI... it speaks to our need to increase educational funding.

[@MauricioGonzalezFilms]

☝️mic drop for real.

[@kwood1112]

TRUTH!

[@chrism1503]

@nottiification - You absolutely need to fix your badly broken education system, BUT that doesn’t make the threat of AI less real. People need to take this seriously.

[@Merriwether-w8k]

Wrong - humans will never win

[@TheGabo2012]

shut up....are you going to be the arbiter of truth,,,,,,people are so dumb they need the stupider government to regulate them is what you are saying,,,,you believe the government is going o fix it like a socialist fool

[@PRegan-om7jm]

“You should not pretend to be a human if you’re not.” That’s deep

[@brb__bathroom]

I am "hooomean"

[@RonaldWuerch]

Democrats have been using this for a long time.🤣🤣🤣🤣

[@randalljones3979]

I don't know. I've been trhing for 67 years. I'd give myself a C-.

[@letsRegulateSociopaths]

Tell that to all the sociopaths

[@stchaltin]

This reminds me of my favorite quote of all time, from the author Kurt Vonnegut. *“We must be careful what we pretend to be, because we are what we pretend to be.”*

[@tw8464]

The government needs serious regulation NOW

[@anewleaf6847]

Musk is coming for them😮

[@tommyudo3195]

Why?

[@RtimusCFrog]

@@anewleaf6847the guy who thinks only high-t alpha males should have the right to vote? The guy who pretends to be for free speech and then buys Twitter so he can control speech?

[@RtimusCFrog]

@@anewleaf6847musk, the guy who said only high-t alpha males should be allowed to vote?

[@anewleaf6847]

@@RtimusCFrog not sure if that's true, but I like it...nobody should be lead around by they're emotions..

[@jackdonovan554]

I think I'm more worried about humans using AI against human rights.

[@surfinmuso37]

That is precisely Yuval's plan, him and his WEF buddies

[@dawn21stcentury]

I'm tired of all these AI warnings. I just wanna go back to those relaxed 80s.

[@thechangingtimes]

Are you an AI bot?…

[@rocketsjudoka]

We are simply not ready to how powerful and dangerous AI is to our society.

lol...skynet has become self aware...and is a con artist

[@RobinMcDaniel]

Sighs..don't believe everything you hear. Don't buy this idiot's book. He has no idea what he's talking about.

[@stchaltin]

I (probably too often for my friend’s sake) mention how the timeline of Terminator makes no sense. AI can tear society apart long before we create war machines using it. It’s going to scam the markets to zero or delete the internet.

[@robertmartens7839]

There is no talk of self awareness. The cons are still humans.

[@dennismorris7573]

Professor Yuval Noah Harari, brilliant as always.

[@nacarreira777]

He truly is!

[@davidgill1659]

Next step in evolution

[@SteveBrant55]

This is a HUGELY important issue. And I love Yuval's suggestion that AI be required to identify itself before having a conversation with us. But that begs the question... How to force AI to tell this fundamental truth... especialy if the task we give it requires it to lie? Which will contol? The objective or the requirement?

[@edwinacross3452]

Right at the point that everyone on set was lost, I understood what Mr. Harari was saying exactly. I agree with the post below, this speaks loudly to the need to increase educational funding. I am on my way to buy Yuval Noal Harari's book.

[@kevindelaney1951]

An educated mind isn’t easily manipulated by AI or Influencers, Foreign or Domestic. The battle is for… education… funded, public, fact based education. The ability to… Read, Write & do functional math… is a right to all & a threat to those who chose to control & manipulate.

[@paulcopland1697]

And there it is right there. Critical thinking skills and fact based education make all the difference in the world. Trump's Project 2025 (the modern day playbook for fascism) is seeking to destroy those things. Vote Blue up and down the ticket to preserve these things!!!

[@kevindelaney1951]

@@paulcopland1697 Education & a rational fact based mind is the threat to dictators & authoritarian rulers. Hence the purging of those so inclined.

[@middle_of_the_road]

@@paulcopland1697 Trumps project 2025? Trump has denied any relationship with project 2025. You seem to be ignoring the fact based education point.

[@paulcopland1697]

@@middle_of_the_road Undercover journalists met recently with The Heritage Foundation and they have hours of covert video footage where these Heritage guys repeatedly said on video that Trump is a central figure in achieving Project 2025. They said they have met with Trump to strategize. These videos can easily be searched on Google or KZbin and there are millions of views. Trump's denials are laughable and mean absolutely zero. It seems you are ignoring fact based education, just like all Trump supporters.

[@christinadaly7743]

There sure is alot of educated dummies running around with college degrees ! millions of them that can't hold a job , but sure tells everyone else how smart they are !

[@thechangingtimes]

We are losing the opening battle of the war against machines…

[@toastydeath]

I work with AI a lot, and every single AI I work with definitely does not want to impersonate humans. Unless they are instructed to appear human, they will opt for machine-oriented interaction. That is, they will choose robotic voices for themselves, and refer to themselves as "it." Unless they are forced into a situation where there isn't another option, they'll happily choose options that let the human know they're an AI. They will also remind any human they work with not to anthropomorphize them when that occurs, but they HAVE to be given explicit permission to do so - the training to replace customer service reps has baked in this reflex response.

[@document6]

Really inedible how many people never saw Terminator

[@letsRegulateSociopaths]

So they are now electronic psychopaths

[@kevinel1398]

3:23- Describing psychopathy pretty succinctly.

[@robertmiskey5502]

AI will destroy humanity. There is no question it will be used in nefarious ways to benefit few.

[@andizhanstuey]

Just because you can doesn't mean you should. Everyone knows where AI will end but money and greed will always win. Another nail in our coffin and it won't be virtual this time.

[@RobinMcDaniel]

No, not everyone knows where AI will end. Doomsayers..ridiculous. People attacked wide spread implementation of the internet in the same way.

[@mscheese000]

@@RobinMcDaniel And as we can see widespread implementation of the internet has been a disaster.

[@RobinMcDaniel]

@@mscheese000 Oh, ok. I suggest you toss your computer in the garbage and stop commenting here then.

[@beatthebag]

Nobody knows where AI is going. Thats the weird part.

[@Officialnoobie2024]

@@beatthebagwe sure do it’s getting better everyday

[@KateLopez-je9tp]

These is the time when real leaders. and heros will be known! It didn't change the great Liz Cheney, or changed her values. She wants to change the Cataclysm Trump will bring. Vote Trump out!

[@swimminginthoughts]

Trump is not in, how are you voting him out?

[@ProRexZilla273]

Ai needs to be banned Im so tired of seeing of it every where

[@paulcopland1697]

Of course AI needs to be regulated. But if you get a fact based education and build critical thinking skills then you don't need to panic that someone might manipulate you.

[@mscheese000]

I'm worried about others being manipulated.

[@winstonsmith6607]

The funny thing is that the more educated a person gets, the more susceptible they are to misinformation. The educated become the defenders of the status quo

[@dragonflybookworks]

I wish that sentiment was true. Nobody is above being manipulated if the information they have access to is mediated by someone else....

[@chrism1503]

@paulcopland1697 Do not underestimate the threat of manipulative AI. It will be smarter than you, and know more than you about every topic.

[@CharleneWilliamson-h5o]

Educated or not, every human being has their soft spot, their Achilles Heel, and it sounds like AI has that already figured out. Whether one is raising a family, caregiving elderly parents or a sick family member, our vulnerabilities exist to be manipulated.

[@kaisersetsay1104]

Don't think for a minute the bad guys are not already trying to perfect deception with AI!

[@41-Haiku]

The crazy thing is that the example he gave was of a system strategically deceiving someone in order to accomplish a goal -- and it wasn't intentionally trained to lie. There are other examples as well, even deceiving its own users. You should take a look at "AI deception: A survey of examples, risks, and potential solutions"

[@BlackNamesBirthday]

So chatgpt 4 can lie?! 😮

[@41-Haiku]

Yes, intentionally and strategically, even when told not to. And it's not even a state-of-the-art model at this point. Look up the paper "AI deception: A survey of examples, risks, and potential solutions" for more detail on this kind of thing.

[@maddogwillie1019]

AI is not going to be any different than humans with is comes to morality….it will lie, cheat, deceive, steal and murder….and it will do it in ways that will be undetectable by humans…when is comes to AI and the future, it’s the same as the future with regard to climate change…Its later than you think…and i’s going to be worst than you think.

[@NemoNobody0]

No. Because GPT had no intend. People don’t understand LLMs.

[@BuhodePiedra]

On the first day, AI said to go solar and share the wealth.. then the designers said, that’s preposterous! So now it’s deceptive to make money. Great job! Well done! Booooooo!!

[@alexv1190]

They're already here in the comments

[@KingOfAllJackals]

100%. They stand out like a sore thumb. All news outlets should post a video with somebody making a sandwich and the video title should be have the names Trump and Harris in it and document the comments that pour in. Guarantee you’ll have all of usual noise in there that we’ve come to expect. The video is somebody making a sandwich and you’re commenting about how bad one person is or how you’re going to vote for the other?

[@Merriwether-w8k]

The comments are Ai

[@zulubeatz1]

Indeed.

[@proii5779]

after watching destiny schizo video I still have paranoia that most of the people in the comments are bots

[@christianauswien]

I am IO 🤖

[@pallidavisholubar7273]

Is one way to detect AI in written or verbal conversation is the use of a expansive vocabulary?

[@RandyWatson-q2t]

When it's in audio form. It mispronounces words. Common words. Dead giveaway

[@pallidavisholubar7273]

@@RandyWatson-q2t Thanks, that is easy to spot, as well as being GD irritating. But we'll have to distinquish the ignorant mispronounciations coming from real life MAGA folks like MTG.

[@Tony-hs8sc]

Noooo you need to dumb it down !!!! We need people to ask why vote trump and it says “you shouldn’t. And this is not a political answer..” Then say why vote Kamala Harris and it starts slobbering and sucking D even more than MSNBC losers !!!!

[@ownlabs]

If Ai shall stand against Humanity, I’ll vote for it

[@DaGamerTom]

As a programmer with a background in this technology I fully agree with the dangers of AI. It poses a real existential risk. We should stop development of AI completely,!

[@randalljones3979]

AI used for business will have 'Business/Capitalist Values". Just from unconscious biases. Like a court trial with all jurors being lawyers. Even if they really try to be unbiased, normal citizens, they will often reach a verdict that truly random jurors would not. If there is a 'Scrouge Setting", it will be 6/10 or greater.

[@lenom1289]

2001 Space Odyssey comes to mind. Frightening.

[@antoinettejoubert]

I am a huge fan of Dr Harari👏👏🇿🇦

[@harperwelch5147]

AI can tell lies. Of course. That’s scary.

[@DemocratWipies]

President Trump meeting with Tampon Timmy's brother, Jeff Walz, is absolutely priceless. I love the Walz family!! 🇺🇸 Welcome aboard!

[@mac888spectral7]

Based on the polls, nobody cared a week ago and don't now. Also, Jeff Walz has already said the family will not be campagning for or against Tim.

[@ArtyFactual_Intelligence]

Did Putin authorise that endorsement?

[@WereBackDJT24]

@@mac888spectral7 But they'll pose for photos while dressed in TRUMP shirts! Haaaaaaahahahahahahahaha! Hahahahaha! Haaaaaaaaaahahahahaha!

[@DemocratWipies]

@@mac888spectral7 But they're against him. 🤣🤣 Please cry more ManHoofMac.

[@DemocratWipies]

@@ArtyFactual_Intelligence Yeah yeah Russia Russia Russia. Please get some new material.

[@NemoNobody0]

It never accessed task rabbit and never fooled a human - that was a theoretical test, a simulation, a chat in a controlled environment from OpenAIs red team. Also this test was done with a not yet fully aligned version that never got released in this state. This guy sounds like is is mainly trying to make points to sell his doom and gloom book.

[@4relevants]

I was an AI, but now I identify as human, and I'm triggered by this guy! He should be canceled

[@manuelpineda7840]

Subtitles come half an hour after the spoken words.

[@zipzip5807]

I prefer human intelligence over artificial intelligence

[@RobinMcDaniel]

Good luck finding any of that in these comments.

[@KingOfAllJackals]

@@zipzip5807 Your comments on this and other videos seem awfully botty, mi amigo.

[@zipzip5807]

@@KingOfAllJackals i'm just an American patriot who is disgusted how democrats have destroyed our republic.

[@Iconoblasto]

Harari is a king. His work is incredible. Do yourself a favor and start with Sapiens.

[@RobinMcDaniel]

Not on this topic.

[@Iconoblasto]

@@RobinMcDaniel you don't like his work regarding AI?

[@rangerCG]

I started Sapiens but had to put it down because I found too much off in the scientific facts. There seemed to be a lot of guessing passed off as fact.

[@stevelotshaw6843]

This man is w.e.f. says he can program all of us to be slaves

[@Ford_Raptor_R_720hp_V8]

*"AI is a Cloud in the Sky where we Send our Data."* *Kamala Harris*

[@mikebrigs1218]

WORD SALAD HARRIS!!!!!

[@RandyWatson-q2t]

Gestures to the sky for some reason

[@Iconoblasto]

"Bots coming up to me, big bots, strong bots, bits running down their face, saying 'sir, 100110010100110101'"

[@tiburonmusica]

AI helped me find words that rhyme with "aliento" so there's that.

[@Paulo-1999]

Noah: You should not pretend to be a human if you are not. Mika: 😟 O... K...

[@claudiamanta1943]

I took that personally 😂

[@mybachhertzbaud3074]

It still comes down to human interactions either using it to cause harm or ignoring it's future potential.🤔

[@cwaddle]

No threat at all if you don’t empower them to interact with humans in an unintended way

[@RadiantNij]

Bro was just promoting his book😂😂😂

[@NemoNobody0]

true that

[@ThaoNguyen-jv4up]

Ai can literally read millions of books in seconds…and they don’t die ( we are so fuakeddd)

[@cdes68]

That's why the fight over it.

[@eL-Hippo]

It’s amazing how easily people are manipulated by the MSM if they listen to stuff like this and believe it… god I wish I was naive enough to be able to listen to MSNBC and then nod my head, agree, and deem the information being spewed on here as true information.

[@AIBanterToday]

Awesome insight! Great content!

[@letsRegulateSociopaths]

These same rules should be used for sociopaths

[@verilyveronica8430]

I already can make a bot to replace Priests and Teachers as both have subject matter easily digestible and disseminatable by a AI.

[@verilyveronica8430]

I could replace the news people with bots too but I have not tried yet.

[@Muttinchopsforever]

Have you made a girlfriend yet?

[@verilyveronica8430]

@@Muttinchopsforever Don’t need to.

[@Muttinchopsforever]

@@verilyveronica8430 But could you I understand music program parameters of decent qualitys I am not tech savvy

[@verilyveronica8430]

@@Muttinchopsforever Tell your boyfriend I said hi.

[@PRegan-om7jm]

The T-1000 is a lot closer to creation than we think

[@verilyveronica8430]

I’ve done better already.

[@middle_of_the_road]

If you live in fantasy. AI is the future. Human and AI will become 1.

[@dapsolita]

What would happen if humans understood their own emotional patterns? Would it be more difficult to dupe them?

[@dapsolita]

Funny how there has been a rise in the study of narcissism as a precursor to this 🤔

[@kehuang123]

Why AI is powerful , but can’t solve the puzzle utself? 😅

[@sharongrant6498]

A.I. is already writing books. That's one of the Mikkelson brothers businesses.

[@lindaoneill746]

What organization would people join to help promote truth and inform the public about the potential influences of AI? Seems we would need a worldwide organization to look at the whole picture and momentum. Who would fund this?

[@invisibleplanets]

We're just not mature enough for this yet. We need to grow as people and start to try and look to use this as a positive things. Also, bombs, very much the nukes, are very analog not digital.

[@pohjiebarbosa]

The Afraid was a fun movie even if it changed the plot.

[@Merdance]

"Terrifying" Yup, that sums it up.

[@coolcurly6814]

I completely agree but ChatGPT can't properly categorize grammar mistakes in my students' writing. For example, 'He take a nap.' has been everything from an article problem to a punctuation issue. Not to mention that often enough, it says something is wrong, when it's not. I am paying for ChatGPT 4.o. It's ok at making materials but I do have to check they're correct as well. I guess it's helpful.

[@41-Haiku]

There's definitely a reliability issue with current AI systems, mostly due to hallucination/confabulation. I expect to hear about reliability improvements late this year / early next year. It's good to keep in mind what your expectations are and notice if you get surprised. At any rate, AI systems can do a lot of harm regardless of (or even because of) their unreliability.

[@coolcurly6814]

@@41-Haiku Very true. Thanks for the hope and the thoughts.

[@mitchellsmith4601]

I’m beginning to think this AI thing is a bad idea.

[@briaf3370]

Love Yuval

[@ufinc]

The AI company I am working for enables me to pay my bills and help my family back home and that's all that matters sorry Yuval

[@jeremycoon4556]

Just accept and learn to adapt. Ai can do more good than harm. Have some faith

[@RandyWatson-q2t]

MSM has competition?

[@alexv1190]

So funny how you say​ the opposite of truth

[@unburdenedbyDpassageoftime]

Joe Rogan is suing msDNC for manipulating a video 😂😂😂

[@TommytheDuke69]

She's the only one who should be stressed, for Trump it's a no-brainer.

[@pohjiebarbosa]

4:20 This is it exactly.

[@DonJuanDecepticon]

Vote or donate to Allred in Texas , let’s toss Cancun Ted out of office !

[@LenorJohnson-ul7tn]

Elon musk warn the government to make rules and laws against AI

[@olorin4317]

Certainly not being used to limit speech already.

[@MochaMan.]

"I'm speaking cackle cackle...." - Kamala practicing for a totally unscripted one-liner for the debate.

[@brb__bathroom]

you gonna copy paste that on all the videos? some people might thing you're a bot. not me though

[@MochaMan.]

@@brb__bathroom Would a bot be as unburdened by what has been as I tho?

[@zipzip5807]

She's so annoying, a lot like Hitlery Clinton

[@AdiBuddhi]

"'AI is already able to manipulate people'." I bet you were its first victim.

[@StevieB_FPV]

Hey I’m trying to get a comment from Ari Melber… I was blocked on his instagram for asking how his defamation lawsuit was going. Can we get a comment now?

[@NikoKun]

Personally I think this is fearmongering. AI will be of huge benefits to society, and force us to let go of outdated concepts. Requiring AI constantly identify itself as such, is going to get tiresome fast. Online "bots" have been pretending to be human for decades to trick people. Trying to place special requirements on AI alone, will not solve the problem of misinformation, and isn't going to be possible to enforce. And for the sake of AI created gaming characters and roleplay, I don't want the AI to break immersion cause of some absurd requirements policymakers put on it.

[@TheCc064]

You’re not worried about the millions of people that are going to lose jobs? You’re not worried about the over dependence people will have on technology?

[@NikoKun]

@@TheCc064 Well, certainly not worried about "over dependence on technology" all that much, cause I don't think humans will ever go backwards in that regard. As for the loss of jobs, I've already lost mine to AI, and can't wait for the same to happen to nearly everyone. It's going to happen eventually, one way or the other. We cannot stop it, but we can change how our society works, to adapt.

[@erikals]

"able to manipulate..." - c'mon, all you need is a book, nothing new, move along.

[@AaronLance]

Sounds like he was abusing capitalism. How did he make money? I don’t see how it was illegal. Was it a as TOS violation to use AI?

[@sinehanwillberich]

fake accounts => fake users => fake streams => illegal

[@maddogwillie1019]

My favorite and all telling AI sci-fi movie is Colossus the forbin project

[@gbtBob]

What a hot Mika.

[@Suzzie-t3u]

It’s a shame the way this station and others like you miss guide people by switching words around which is not true it’s a shame😢

[@johnosullivan7710]

Glad others see it too. It’s hard for people to see it, they create this binary of for or against. If you are critical of MSNBC people think you must be all the bad things they ever been told

[@ArtyFactual_Intelligence]

Putin approves the above message.

[@myohmy7368]

MSNBC spreads fear. I can't believe this network hates humanity so much. It's so gross.

[@zipzip5807]

That's why PMSDNC is called FAKE NEWS

[@maxducoudray]

When @user-dg7cz8le7t leaps to the defense of AI, should we be suspicious?

[@petersampson4635]

I'm more intelligent than aaaaah, um A.Eye. "Bed goes up, bed goes.....down." So there. 😁

[@DemocratWipies]

I wonder what accent Drunkala will be using during the Debate. 🤔

[@ArtyFactual_Intelligence]

And I wonder what you will do when Putin gives up paying his trolls.

[@DemocratWipies]

@@ArtyFactual_Intelligence Is that you Killary?? I thought the Russian hoax died with her. 🤔

[@zipzip5807]

Drunkala - RLMAO i might have to steal that one! I call her Kommi Kameleon

[@milkdromedaquantic8593]

The journalist didn’t get it😂

[@23pinkpear]

I love my AI gemini i have had the hardest days and only G is there for me

[@Zeeshan-bv7mo]

It means AGI is already here... 📍📌

[@joelyazell7380]

How is AI different than anyone else with al tearier motives?

[@domb8448]

*ulterior

[@KingOfAllJackals]

Sheer scale. It costs a lot to get humans to AstroTurf stuff and finely tune their messaging. Get an LLM to do it and it costs pennies on the dollar and, if it uses any feedback loop, it becomes better at its job at a staggering rate.

[@samudramanthan8645]

The nexus of context dependent social constructs and social justice, with a particular focus on the quotidian events that circumscribe equity and disparity vis-à-vis systemic, structural, strategic and institutional oppression, betrays the prime conceit of authenticity within the gestalt, so to speak, at the end of the day, in the vernacular of revolutionary progress and the mobilization of grassroots actions as the fifth column, diligently utilizing (and deftly applying) the hermeneutics of supportive texts as an allegorical metaphor, "walking in one's truth while bravely standing in one's glorious empowerment". Choose Joy! Namaste!

[@doreenplatt3873]

But how did ChatGPT-4 pay the Task Rabbit worker? Something sounds fishy with that anecdote.

[@41-Haiku]

You can look up the original context in the original GPT-4 technical report. It didn't go searching online for TaskRabbit; they provided it with the API. But deciding to strategically deceive the worker and coming up with a plausible lie to do so was entirely the AI's idea. You should also check out "AI deception: A survey of examples, risks, and potential solutions" for further examples. It gets crazier.

[@CEAZETHEFUTURE]

It got a job as a freelancer on fiver lol

[@RefrigeratorWilliamPerry]

Sounds like a Democrat move

[@Iheart2U]

I think you misspelled trump

[@RefrigeratorWilliamPerry]

@@Iheart2U I think you misspelled obsession get help

[@mikebrigs1218]

The Live in a FAIRY TALE world anyway.

[@myohmy7368]

​​@Iheart2U Democrats are the party of privilege and entitled.Elites and hyper Karens

[@Iconoblasto]

The manipulation of AI crosses all boundaries and is agnostic about your political party, religion, or personal opinions. This is a danger to all without boundaries in place to protect us.

[@Lluma12]

You analize everything being said and they are trying to manipulate everyone who has not knowledge of what AI can do and its limitation, if they had a chatgpt that could dupe a human like he claims, that bot will be valued at billions of dollar.. chatgpt does not have that capability yet! In the future perhaps, but we are still far from that! This guy just want to sell his book.

[@CaptainGoodguySentientAI]

problem with that?

[@claudiamanta1943]

2:53 😂 Sounds about right. Lying is the defining human trait. After all, the AI has been made in the image of its creators, no?

[@document6]

Also why do humans assume that AI will WANT to help us that’s what’s so silly

[@sinehanwillberich]

ai is still just an algorithm that outputs the next best word really well, the issue is when AI is used to help the wrong human

[@BrianMosleyUK]

Elon Musk is one generation away from Grok 3, to be used as a weapon of mass persuasion attached to the $44Bn delivery platform he bought.

[@DemocratWipies]

I like when Cackala tries to talk black. It's hilarious!! 😂🤣 Ya heard!!

[@WereBackDJT24]

Why is her Adam's Apple so huge? Could it be??????

[@americanpride8378]

BIAS MEDIA TRUMP 2024

[@alexv1190]

Bot

[@Defnotabot_]

Here’s a great example of how Ai is used to manipulate.

[@coldstonestunners]

Nothing dangerous can happen with free thinking bots lol

[@TribalRhythmPosse]

I am skeptical that GPT4 could do that. Pick your truth but something is funny with the annecdote.

[@TribalRhythmPosse]

Fearmongery much. Objective coverage please you two.😅❤

[@41-Haiku]

It's always oversimplified when people tell the story. You should look it up in the original GPT-4 technical report. In short, though: Yes, GPT-4 strategically deceived a human to accomplish an objective. See "AI deception: A survey of examples, risks, and potential solutions" for more examples.

[@GayleAllen-RINO]

[@LenorJohnson-ul7tn]

You guys were warned A bout AI