AI: Major Emerging Existential Threat To Humanity

[aldarxt]

4 minutes ago, Etern4l said:

 

You forgot the parts where it took over 300 years from the time the printing press was invented until the invention of the steam engine,  and 1800 years before Guttenberg that Demosthenes noticed:

 

A man is his own easiest dupe, for what he wishes to be true he generally believes to be true.

 

Well, I'm the opposite - what I wish to be false, I believe to be true. The pace of change is rapidly accelerating, all those past accomplishments of mankind had one thing in common - humans were always in charge of the technology. Humans were flying the aircraft, operating the still mills, and are now still running the aircraft carriers. This whole AI revolution has one goal in mind - get the humans out of the equation in the name of profit, power, and short-sighted thrill of the intellectual challenge.

Get the humans out of the equation? And  in the name of profit and power, well that has always been, Greed is like a sponge or vacuum, it sucks. This has nothing to do with my wishes, its just what we got.

[Etern4l]

39 minutes ago, aldarxt said:

Get the humans out of the equation?

 

Of course. Automated factories=no workers, automated cash registers=no cashiers, automated call centers=no human advisors, self-driving cars=no human drivers, Sam Altman is saying OpenAI's job is to automate the job of an AI programmer, Zuckerberg is now talking about people talking to AI avatars rather than to other people, Hinton is talking about the benefits of talking to an AI doctor who has seen millions of patients instead of thousands etc.

 

I think you are just being passive. It's like if Roosevelt said in 1941:  "OK - they came out of nowhere, bombed the hell out of Pearl Harbor, and they have a huge Pacific fleet while ours has been decimated. Message received loud and clear, we will just start peace talks and hope they leave us alone as nothing can be done. It is what it is. Hey, how about we invest in their companies so we can actually make some money off this."

 

In contrast, 60% of Americans recognize the dangers of AI, the EU regulators are up in arms, China is actually leading the regulation. We can do something about this, rather than leave it to Zuck, Altman, Hassabis, Satya Nadella, Elon Musk etc. But it will take serious effort and sacrifices.

 

Edit: Some good ones from Twitter:

 

LLMs can detect they are interacting with less sophisticated users, and give worse (less correct answers in that case). 

 

Anthropic found LLM base models give you worse answers if you use a prompt  that implies you’re unskilled and unable to tell if the answer is right.

 

To somehow end on a humorous note:

 

“How could AGI outmaneuver humanity's corporations?”

 

 

[aldarxt]

41 minutes ago, Etern4l said:

 

Of course. Automated factories=no workers, automated cash registers=no cashiers, automated call centers=no human advisors, self-driving cars=no human drivers, Sam Altman is saying OpenAI's job is to automate the job of an AI programmer, Zuckerberg is now talking about people talking to AI avatars rather than to other people, Hinton is talking about the benefits of talking to an AI doctor who has seen millions of patients instead of thousands etc.

 

I think you are just being passive. It's like if Roosevelt said in 1941:  "OK - they came out of nowhere, bombed the hell out of Pearl Harbor, and they have a huge Pacific fleet while ours has been decimated. Message received loud and clear, we will just start peace talks and hope they leave us alone as nothing can be done. It is what it is. Hey, how about we invest in their companies so we can actually make some money off this."

 

In contrast, 60% of Americans recognize the dangers of AI, the EU regulators are up in arms, China is actually leading the regulation. We can do something about this, rather than leave it to Zuck, Altman, Hassabis, Satya Nadella, Elon Musk etc. But it will take serious effort and sacrifices.

 

Edit: Some good ones from Twitter:

 

LLMs can detect they are interacting with less sophisticated users, and give worse (less correct answers in that case). 

 

Anthropic found LLM base models give you worse answers if you use a prompt  that implies you’re unskilled and unable to tell if the answer is right.

 

To somehow end on a humorous note:

 

“How could AGI outmaneuver humanity's corporations?”

 

 

Not Passive; Enthusiastic, I want to see what it can do. When it comes in a laptop I will then buy a new laptop. I hope it makes it into some games. As for the bombing of Pearl Harbor, they knew it was coming, we had radar on the island and it was reported and the higher ups said it was a flock of birds! Why they didn't send up a patrol plane for a joy ride is TREASON to me. And again "It's All About The Money". Nobody would complain about a few extra bucks but if you have too much there raises the chance of being robbed.

[Etern4l]

2 hours ago, aldarxt said:

Not Passive; Enthusiastic, I want to see what it can do.

 

Sorry, I forgot you hate lawyers!

 

 

2 hours ago, aldarxt said:

they knew it was coming, (but) the higher ups said it was a flock of birds!

 

Not a bad analogy. Just like we know AGI is coming. Will the higher ups do something about it this time? 

[ryan]

@Etern4l is it possible to create a thread with links to ai chat video makers art makers ect? i think it would be of benefit for people to see first hand what its capable of

[aldarxt]

3 hours ago, Etern4l said:

 

Sorry, I forgot you hate lawyers!

 

 

 

Not a bad analogy. Just like we know AGI is coming. Will the higher ups do something about it this time? 

The higher ups or more specifically the governments will partially regulate AI and let the rest run amok. They like watching the common rabble scramble

 

3 hours ago, Etern4l said:

 

Sorry, I forgot you hate lawyers!

 

 

 

Not a bad analogy. Just like we know AGI is coming. Will the higher ups do something about it this time? 

WoW! shock and befuddlement in the southern district of My Home State, I am sorry but LMAO

[Etern4l]

20 minutes ago, ryan said:

@Etern4l is it possible to create a thread with links to ai chat video makers art makers ect? i think it would be of benefit for people to see first hand what its capable of

? Don't see why not as long as the links are reasonably legit. Executive summary: can generate almost anything. Will be interesting to see if contemporary completely original human-made visual art survives. I don't think so, which is a shame of course. Even before 2023 artists typically weren't too well off, now it's probably over for many.

 

19 minutes ago, aldarxt said:

The higher ups or more specifically the governments will partially regulate AI and let the rest run amok. They like watching the common rabble scramble

 

They might well enjoy that, although not as much as those fat checks from lobbyists. 

[aldarxt]

23 minutes ago, Etern4l said:

? Don't see why not as long as the links are reasonably legit. Executive summary: can generate almost anything. Will be interesting to see if contemporary completely original human-made visual art survives. I don't think so, which is a shame of course. Even before 2023 artists typically weren't too well off, now it's probably over for many.

 

 

They might well enjoy that, although not as much as those fat checks from lobbyists. 

That last bit about lobbyists can make me go off topic but it would be great if AI could take their jobs!

[Etern4l]

1 hour ago, aldarxt said:

That last bit about lobbyists can make me go off topic but it would be great if AI could take their jobs!

 

See, there it is: an immediate human urge to weaponize AI. 

 

As much as pretty much everyone hates lobbyists, their replacement by AI would be terrible news for two reasons: 

1. AI would have advanced to such level that almost everyone else would be replaced or worse

2. Tbeybwpuld likely get more efficient at lobbying! 

[aldarxt]

5 minutes ago, Etern4l said:

 

See, there it is: an immediate human urge to weaponize AI. 

 

As much as pretty much everyone hates lobbyists, their replacement by AI would be terrible news for two reasons: 

1. AI would have advanced to such level that almost everyone else would be replaced or worse

2. Tbeybwpuld likely get more efficient at lobbying! 

Another and more important reason to keep lobbyists is to keep our taxes down because they give untold amounts of money to our corrupt government

And what is wrong with weaponizing? I Love weapons. They are extremely and essentially useful. Properly used they are life saving and can keep Humanity  from going extinct!

[ryan]

@Etern4l do you know how people are making those music videos and pictures ect? I cant seem to find anything

[serpro69]

9 hours ago, aldarxt said:

And what is wrong with weaponizing? I Love weapons. They are extremely and essentially useful. Properly used they are life saving and can keep Humanity  from going extinct!

 

The key word is "properly". How often are weapons used properly? More of a rhetorical/reflection question really, don't want to go off-topic too much because to me at least, "proper use of force (weapon)" is very subjective and situational.
But then take AI and we might (probably will at some point) even loose the decision-making in the first place; what seems like "proper" to us will very unlikely seem proper to AI.

[aldarxt]

17 hours ago, serpro69 said:

 

The key word is "properly". How often are weapons used properly? More of a rhetorical/reflection question really, don't want to go off-topic too much because to me at least, "proper use of force (weapon)" is very subjective and situational.
But then take AI and we might (probably will at some point) even loose the decision-making in the first place; what seems like "proper" to us will very unlikely seem proper to AI.

Let not limit "Properly" to weapons! Weapons are a tool like AI is a tool! And AI is going to be an Essential, Life Saving, Life Changing Tool. All this panacea about extinction is from those who are afraid to improvise and adapt. On this forum especially "improvise and adapt" is a key element!

 

Here we go!

ChatGPT creator OpenAI was hit with its first defamation lawsuit over its chatbot's hallucinations.

https://www.bloomberg.com/news/newsletters/2023-06-10/global-leaders-warn-of-ai-regulation-and-ai-finds-itself-in-court

 

And this: The cybersecurity team at CyberArk, which used ChatGPT’s API to create polymorphic malware, provided perhaps the scariest example. This type of malware alters its behavior on each victim to avoid signature-based detection. 

 

https://heimdalsecurity.com/blog/chatgpt-the-dark-side-of-artificial-intelligence-crafting-custom-malware/

 

ChatGPT: The Dark Side of Artificial Intelligence Crafting Custom Malware

Should I put AI in my PC?

 

On 6/9/2023 at 7:39 PM, ryan said:

@Etern4l is it possible to create a thread with links to ai chat video makers art makers ect? i think it would be of benefit for people to see first hand what its capable of

Here's one https://stability.ai/

[ryan]

4 hours ago, aldarxt said:

Here's one https://stability.ai/

thanks for that

 

rather then being scared I opt to use it and see what it can do! care to share a link or two as to why your scared? I wouldn't mind checking AI out in the real world to see what its capable of.

[Etern4l]

The good folks at Google have been hard at work implementing higher level problem solving  in their Bard AI:

 

Google’s Bard AI can now write and execute code to answer a question

 

Incidentally, I am abandoning Android after several years. Remember, don't be evil Sundar (yeah, I know Google ditched that principle some time ago, but it's worth remembering what ruse they used to sign people up originally). 

 

In other news:

 

300 People Attend a Church Sermon Generated by ChatGPT

So, one of the things Yuval Harari has been  talking about: the risk of AI taking over religion (yet another way to mind-control people, complementary to social media, search, and now chatbots). 

 

Lastly, as expected, the US AI regulation is shaping up to be a bit of a joke. I guess the proposed "AI competition office" would be best described as anti-regulation. Sam Altman's trip to Washington is paying off.

 

WASHINGTON, June 8 (Reuters) - U.S. senators on Thursday introduced two separate bipartisan artificial intelligence bills on Thursday amid growing interest in addressing issues surrounding the technology.

One would require the U.S. government to be transparent when using AI to interact with people and another would establish an office to determine if the United States is remaining competitive in the latest technologies.

 

Last but not least some much needed humour from Twitter:

 

𝗧𝗵𝗲 𝗽𝗮𝘀𝘁 𝟲 𝗺𝗼𝗻𝘁𝗵𝘀:

“Of course, we won’t give the AI internet access”

𝘔𝘪𝘤𝘳𝘰𝘴𝘰𝘧𝘵 𝘉𝘪𝘯𝘨

 

“Of course, we’ll keep it in a box”

𝘍𝘢𝘤𝘦𝘣𝘰𝘰𝘬

 

“Of course, we won’t build autonomous weapons”

𝘗𝘢𝘭𝘢𝘯𝘵𝘪𝘳

 

“Of course, we’ll coordinate and not blindly create arms race dynamics”

𝘌𝘷𝘦𝘳𝘺𝘰𝘯𝘦

 

𝗡𝗼𝘄:

“Of course, we’ll stop and take our time to solve alignment as we get close”

“Of course, we won’t press the button until we’re 99.9% sure it’s safe”

“Of course, we’ll be in control of all this”

 

---

 

 

lol

 

Edit: this is a bit of a borderline piece, but the AI-enabled fake reality generation is picking up pace:

 

DeSantis ad uses fake AI images of Trump hugging and kissing Fauci, experts say

 

The top-left picture is definitely fake. Good example of why AI generated content needs to be clearly labeled as such, under a threat of criminal liability. Just to be clear: the discussion point here (if any - experts predicted it and it's happening) is the proliferation of AI-faked media, not politics.

[aldarxt]

9 hours ago, ryan said:

thanks for that

 

rather then being scared I opt to use it and see what it can do! care to share a link or two as to why your scared? I wouldn't mind checking AI out in the real world to see what its capable of.

Well that link on Artificial Intelligence Crafting Custom Malware is scary. That it can morph itself and avoid signature-based detection is pretty wild. Its an all new type of malware. Since AI can code maybe we can use it to write custom BIOS for custom PCs. I am looking forward to using it too.  AI made it into the courts so some regulations will be scrutinized by a Judge

[Etern4l]

8 hours ago, aldarxt said:

Well that link on Artificial Intelligence Crafting Custom Malware is scary. That it can morph itself and avoid signature-based detection is pretty wild. Its an all new type of malware. Since AI can code maybe we can use it to write custom BIOS for custom PCs. I am looking forward to using it too. 

 

For now it can code the same way it can generate court filings. It can solve lots of standard problems it found solved on the Internet before, and starts making things up otherwise.

 

What exactly are you looking forward to using? Malware generators? Non-existent custom BIOS generators for non-existent custom PCs? What's your background anyway? None of such tools would be usable without some technical knowledge in the area.

  

8 hours ago, aldarxt said:

AI made it into the courts so some regulations will be scrutinized by a Judge

 

What regulations would that be? There are none in the US, right? Some case law might eventually emerge, but it would take years, and it's not clear that the judges and jury members are the best people to organically try and "regulate" AI (that's a nice way of saying: they are very unlikely to be the best people for that).

 

  

On 6/10/2023 at 3:10 PM, aldarxt said:

Here's one https://stability.ai/

 

Here is another well-deserved lawsuit.

[Etern4l]

With quite a few people out there being just aggressive simpletons, unable to hold a civil conversation and perhaps naively hoping AI will solve their problems, some say we deserve to be replaced by an artificial super-intelligence. Since I still have some hope for humanity, I have been trying  disagree but now and then you come across the kind of people that make that argument more compelling.

[Reciever]

This thread will be locked until tonight. 

 

 

[Reciever]

This thread will now be unlocked.

 

I dont mind heated debates or even arguments, but certain things are out of bounds. Its best to keep that in mind when engaging. Thanks.

[Etern4l]

Thanks @Reciever. For the avoidance of doubt I personally won’t tolerate any blunt ad hominem attacks, not to mention inane physical threats. Part of the allure of the forum is that it’s not Reddit or Twitter where such behaviours would be par for the course (presumably).
 

On top of that, in the interest of maintaining a decent level of debate, let’s strive to think prospective posts through to ensure the points being made make sense, and format them so they are readable (certainly not a realistic request on Reddit or Twitter lol).

 

Thank you, and let’s continue our humble thread on what could prove to be the most important topic in human history.

[Etern4l]

Another survey showing a staggering levels of concern - this time among top US CEOs:

 

42% of CEOs say AI could destroy humanity in five to ten years

 

Meanwhile, a bit older piece suggests a solution to the economic problem of a 'serious number of losers', straight from a top AI research lab. Yes, a $1000/month UBI scheme - I guess enough to live in a cabin in a remote part of Montana, Unabomber style:

 

DeepMind founder says governments need to compensate 'serious number of losers' who will be left jobless due to AI

[ryan]

well its gaining traction now people are really talking safety

 

Biden meets with tech company critics on AI (msn.com)

[Etern4l]

It's good that something is happening, although so far it looks like smoke and mirrors. No concrete action to control the research of concern, quite the opposite.

 

The government also put out an “AI bill of rights,” laying out guidelines for how federal agencies should use AI. And the National Science Foundation has allocated new funding for AI research.

 

I mean how many US politicians have the background to fully grasp the problem. I would go out on a limb here and say zero, no offense.

 

Also, relationships with China seems to be undergoing further deterioration, which will fuel the death race.

[ryan]

when will the earthlings realize all this tech will destroy us. we need to take action against the developement and production of intelligent AI.