AI: Major Emerging Existential Threat To Humanity

[Etern4l]

We have a "limited free will". We can attempt to choose among the ideas/potential actions that "come to mind", i.e. what the subconscious part of the brain suggests to us. The key question is: what does come to one's mind and why? This is most likely based on the more recent input, now controlled by big tech to a dangerous degree, but then what we actually do is heavily dependent on our fairly primitive reward-motivation pathway which also gets hacked and hijacked by technology. All worked great back in on the savanna 100,000 years ago, not so great today, in the era of AI-controlled, and soon generated, (anti-)"social" media.

[ryan]

I think the problem is an overlap in progress. We are evolving slow and technology is evolving fast. If we don't mitigate the gap we really are doomed. also I'd like to add what I thouhgt you meant was really it. I'm fairy perceptive but I just had to be sure, are you right? Afraid so, I'm positive this is truth and we are in trouble if we don't act not react

[serpro69]

On 5/20/2023 at 1:34 PM, Etern4l said:

Here is another clip from an interview with Yuval Noah Harari, a historian, futurist and renowned author (recommend watching the whole thing):

Thanks for sharing, watched a couple minutes and thinking I'm going to start from the beginning - looks quite interesting.

[Etern4l]

From 11:00 - Elon is working on AGI and ultimately building 16 billion Optimus humanoid robots. What could go wrong... 

 

 

https://youtu.be/D_8WShgZLQM

[serpro69]

On 5/18/2023 at 11:28 PM, ryan said:

it's threatening us how again?

 

Here's a quite simple and quick explanation from one of AI pioneers

 

 

[cucubits]

On 5/22/2023 at 7:31 AM, Etern4l said:

From 11:00 - Elon is working on AGI and ultimately building 16 billion Optimus humanoid robots. What could go wrong... 

 

 

https://youtu.be/D_8WShgZLQM

 

I lowkey hope he's going to name them "t-100"

 

[Etern4l]

14 hours ago, cucubits said:

 

I lowkey hope he's going to name them "t-100"

 

 

He already called them Optimus or something. He has been on a cunning campaign of AI safety virtue signalling for dummies, so surely will continue to avoid any obvious PR issues as he continues on his Iron Man-esque quest for world domination. 

[ryan]

On 5/22/2023 at 8:31 AM, Etern4l said:

From 11:00 - Elon is working on AGI and ultimately building 16 billion Optimus humanoid robots. What could go wrong... 

are u joking

[Etern4l]

8 minutes ago, ryan said:

are u joking

 

Did you watch the Tesla event video? He said he ultimately expects one or two of those robots per human inhabitant of the planet. Billions. Obviously part of it is marketing BS aimed at getting dumb money to part with their cash, but he wasn't obviously joking. 

[ryan]

Search hyperbole.

 

I'm looking to build a time machine and take over the world...now wait for it..like this is getting silly

[Etern4l]

Eating Disorder Helpline Fires Staff, Transitions to Chatbot After Unionization

What's interesting about the story is that the bot is supposedly not GPT-based, so even simpler technology can well wreak havoc on the labour market. So which human jobs can't technology take? Well, let's cross providing advice and talking empathically to people in need off the list.

 

This example is about greed pure and simple, and here AI is a technobabble excuse to ... people and siphon resources to those at the top of the heap (running the "non-profit" in this case).

 

  

On 5/22/2023 at 4:44 PM, serpro69 said:

 

Here's a quite simple and quick explanation from one of AI pioneers

 

The problem is how to effectively communicate this to a very broad audience, including people who are non-technical, poorly educated etc. 40% of the US population fails to grasp these issues according to a recent survey. You'd basically need an anti-AI Trump, who probably wouldn't speak against AI himself, given that it/ML played a large part in the online side of the effort to get him elected. Interestingly though, there is a lot of understanding on that side of the electorate already.

People who are already feeling the heat of AI are more aware. If you are a trucker, you don't need convincing, since you are understandably alerted by the efforts to completely replace you with autonomous driving.

 

If you are a border force officer, you may be having a bit of laugh at this one:

 

Passport e-gates at UK airports down

 

Basically it's a failure of an AI system deciding whom to let into the country....

[Etern4l]

Here we go:

 

 

 

On the lighter topics of destruction of visual art and the generation of post-truth world:

 

 

 

 

[ryan]

must watch video below

 

 

[Etern4l]

14 minutes ago, ryan said:

must watch video below

 

 

 

I saw that, thought it left something to be desired. Generally commercial mainstream media, especially on the right side or connected to big tech like CNBC, tends to downplay AI risks. "We need to consider things like universal income to support people in the 'transition'". Transition to what lol? Prompt engineering? Any idiot can do that. Where is the money for this universal income going to come from?

 

Here is a somewhat funny story though - a lawyer submitted a court filing copied and pasted from ChatGPT, with made up case references. Of course, all those quirks will be ironed out within a year or two max.

 

Lawyer 'Greatly Regrets' Relying on ChatGPT After Filing Motion Citing Six Non-Existent Cases

[Etern4l]

People are waking up. Some protesters in London at the venue where Altman and Hassabis I believe were speaking:

 

Artificial intelligence could lead to extinction, experts warn

Meanwhile industry leaders hard at work trying to create the very existential risk, i.e. AGIs, have signed a declaration of "concern" over the threat it would pose:

OpenAI execs warn of “risk of extinction” from artificial intelligence in new open letter

Statement on AI Risk

 

That gall of those people, but ultimately unbelievable human stupidity in failing to stop them right away.

 

Meanwhile, Mad Jensen unveiled a new exaflop supercomputer, which could well be the machine that is used to actually create an AGI. He said it will be used for "exploratory research" (translation: AGI research) by the usual gang of bad guys: Google, Meta, Microsoft/OpenAI obviously, and Nvidia themselves.

 

 

[aldarxt]

Anybody remember "Henny Penny"?

[Etern4l]

Err, anyone remembers the story of Cassandra?

 

All those senior AI figures and AI lab bosses are coming out with warnings for a reason. It's unprecedented. Did tobacco firm bosses ever issue any warnings about lung cancer without anyone forcing them to do so? It doesn't happen in capitalism. 

 

 

 

[serpro69]

Sam Altman did a tour last week around Europe discussing AI Act with government officials , and his immediate reaction was "do not over-regulate or OpenAI will leave Europe" ( https://www.reuters.com/technology/openai-may-leave-eu-if-regulations-bite-ceo-2023-05-24/ ) , only to shit himself when his threats backfired ( https://www.reuters.com/technology/openai-ceos-threat-quit-eu-draws-lawmaker-backlash-2023-05-25/ ) and promptly say on the next day "we're not planning to go anywhere"  ( https://www.reuters.com/technology/openai-has-no-plans-leave-europe-ceo-2023-05-26/ )

Don't you just love it when corporations try to threat governments 

[Etern4l]

AI-controlled US military drone ‘kills’ its operator in simulated test

In the exercise, AI used creative strategies to set itself free from control, such as destroying comms tower and killing the operator... 

 

Meanwhile, Joshua Bengio, one of the godfathers of AI, joins the chorus of experts warning us about the existential and other dangers of AI:

 

 

[ryan]

they denied it ever happened. usually they tell the truth. its bs bro

[Etern4l]

4 hours ago, ryan said:

they denied it ever happened. usually they tell the truth. its bs bro

 

Lol, the military usually tell the truth about presumably classified highly controversial operations, don't they? Dude please change the avatar, for Albert's memory's sake...

 

The guy said too much in good faith (wanted to highlight the unexpected behaviour AI is capable of), received a reprimand, backtracked, and a spokesman issued a lame statement of denial nobody with at least half a brain would believe. 

 

Furthermore, the behaviour he described is exactly what I would expect in absence of sufficient guardrails, especially in some early testing. 

 

A side-point here is that the US military are working on autonomous drones, which has been known for a while. 

[Reciever]

9 minutes ago, ryan said:

I don't know whats wrong with you man, but I fear you need help. AI won't kill us common stupidity will

 

7 pages of posts and not 1 sentence of fact!

 

lol. check in or check out bud

If you can't engage in a constructive manner, then it's time for you to check out. 

[serpro69]

10 hours ago, Etern4l said:

In the exercise, AI used creative strategies to set itself free from control, such as destroying comms tower and killing the operator...

 

Quote

AI had used “highly unexpected strategies to achieve its goal”

 

Lol, how exactly is it unexpected? AI will do anything to get more points ... it's at the core of its training.

  

2 hours ago, Etern4l said:

Furthermore, the behaviour he described is exactly what I would expect in absence of sufficient guardrails, especially in some early testing.

 

They put in guardrails later, explicitly prohibiting it to kill the operator - didn't help much, the drone went for the comms...

 

 

On a slightly different note, https://www.washingtonpost.com/technology/2023/06/02/ai-taking-jobs/ - some stories of people who lost jobs to ChatGPT, despite companies themselves admitting that the quality is worse (but of course cost savings will trump quality 9 times out of 10)

[Aaron44126]

12 hours ago, Etern4l said:

In the exercise, AI used creative strategies to set itself free from control, such as destroying comms tower and killing the operator...

 

This appears to be a bogus story.  (Indeed, it seems to be something more out of science fiction and it is easy to poke holes in.)

 

Some AI stories/videos I've run across recently.

AI in education (...Actually pretty cool, if it can be pulled off.)

AI researcher on tripping up ChatGPT 4 with simple queries

(Despite how "smart" and "confident" it sounds, ChatGPT is basically spitting out amalgamations of its training data – that's exactly how AI neural networks work – and not actually "thinking" like a person does.  Throwing more training data at it won't necessarily fix this because "common sense" isn't something that is spelled out on the Internet for a machine to pick up and "learn".)

Geoffrey Hinton (long-time researcher, at Google from 2013 until pretty recently) warning about dangers of AI

 

Right now, I'm less worried about "death and destruction" caused by AI and more worried about it displacing real human jobs in certain sectors, and also stirring up the misinformation scene even more than it is already.  These are real problems happening now and likely to get significantly worse over the next few years.  There doesn't seem to be anything in the works to seriously address either problem, either.

[Etern4l]

5 hours ago, serpro69 said:

Lol, how exactly is it unexpected? AI will do anything to get more points ... it's at the core of its training.

 

You can imagine senior military personnel being surprised. If you put a human test pilot in an aircraft, virtually in no circumstances would you expect the result to be deliberate friendly fire, and of course the contractor themselves clearly wouldn't have advertised that as a possibility (even in a simulation). Yet, of course it could be a perfectly rational thing for the AI to do. You can try to work around this, but fundamentally it's an alien intelligence. It has no morals or other human values, no empathy, and ultimately limited what we would call "human common sense" at best.

 

I know it's a bit of a distraction, but being a bit of a movie buff, I am failing to resist the urge to share a classic scene which comes to mind here lol

 

 

 

5 hours ago, serpro69 said:

On a slightly different note, https://www.washingtonpost.com/technology/2023/06/02/ai-taking-jobs/ - some stories of people who lost jobs to ChatGPT, despite companies themselves admitting that the quality is worse (but of course cost savings will trump quality 9 times out of 10)

 

Yeah, this might soon warrant a separate thread. Looking at the comments, people try to console themselves with the imperfections of ChatGPT etc. (probably based on the initial/free version), but of course we should expect these to be ironed out very soon indeed, if they haven't been already. People will try to run away to manual labour, but there are only so many dog walkers, plumbers, and HVAC engineers the world needs (and Elon is hard at work trying to progress robotics and commence an assault on that front). A good point was made by Yuval Harari in one of the videos earlier: only a few years of ca. 20% unemployment was enough for a well known populist movement to grab power, which lead to WWII.

 

Edit: in response to @Aaron44126, welcome back to the thread :)

  

3 hours ago, Aaron44126 said:

 

This appears to be a bogus story.  (Indeed, it seems to be something more out of science fiction and it is easy to poke holes in.)

 

As mentioned earlier, I don't think the story is bogus. An army colonel stood in front of a sizable audience and shared it, and it makes perfect sense - he was just describing some behaviour observed during training/testing. Would the military have confirmed it? Of course not, and now it's PR damage control, especially in light of the talk about AI regulation.

 

I mean the current progress in AI would be considered sci-fi not long ago, certainly 10-20 years back.  What holes do you see?

  

3 hours ago, Aaron44126 said:

AI in education (...Actually pretty cool, if it can be pulled off.)

 

Better than no education for sure, but potentially really dangerous. One Valley company in control of the system educating billions in third-world countries. Hmm.

 

If anything, I am personally more excited by the potential of AI to help improve healthcare, e.g. diagnostics

 

https://www.theguardian.com/society/2023/apr/30/artificial-intelligence-tool-identify-cancer-ai

 

This is narrow AI tech which is relatively harmless apart from maybe through some impact on radiologists as a profession (hopefully there will always be people sanity checking stuff at least).

  

3 hours ago, Aaron44126 said:

AI researcher on tripping up ChatGPT 4 with simple queries

(Despite how "smart" and "confident" it sounds, ChatGPT is basically spitting out amalgamations of its training data – that's exactly how AI neural networks work – and not actually "thinking" like a person does.  Throwing more training data at it won't necessarily fix this because "common sense" isn't something that is spelled out on the Internet for a machine to pick up and "learn".)

 

Haven't watched end-end, but couple of problems there. The issue is the rate of improvement in capability, including reasoning. 

Second, I tried a few of those with 3.5 and was able to get the right answers with follow up queries. I mean it's well known that ChatGPT is still a bit of an idiot savant. Check out the story I posted earlier about a law firm getting sanctioned by the court for submitting a ChatGPT-generated filing containing made-up case references.

 

Furthermore, comparisons to human thinking are a bit pointless because:

a) We don't really understand how our brain works

b) AI will almost necessarily be "thinking" very differently to us - how it does it doesn't really matter in the end if it continues to gain function at this staggering pace

 

Remember that folks appearing in TED talks are usually advertising something (in that case an alternative approach to AI).

 

  

3 hours ago, Aaron44126 said:

Right now, I'm less worried about "death and destruction" caused by AI and more worried about it displacing real human jobs in certain sectors, and also stirring up the misinformation scene even more.  These are real problems happening now and likely to get significantly worse over the next few years.

 

Yes, so as far as the public knowledge goes, we are not facing an immediate and direct existential threat from AI, however, we are clearly on course. Hopefully there is still enough time to prevent it, but probably not much.

 

However, as mentioned above, the immediate issues such as the impending significant unemployment (which is almost a foregone conclusion now) or breakdown of democracy can very well lead to death and destruction, vide 1930s Europe and then the world.