Is chat GPT a danger to humanity?

General chat topics, anything and everything you want or need to discuss
User avatar
Horse
Posts: 11558
Joined: Sun Mar 15, 2020 11:30 am
Location: Always sunny southern England
Has thanked: 6196 times
Been thanked: 5088 times

Re: Is chat GPT a danger to humanity?

Post by Horse »

Horse wrote: Fri Jun 02, 2023 8:02 am
Count Steer wrote: Fri Jun 02, 2023 7:44 am 'I'm sorry Dave. I'm afraid I can't do that'.

https://www.theguardian.com/us-news/20 ... p_Other
Interesting that:

In a statement to Insider, Air Force spokesperson Ann Stefanek denied that any such simulation has taken place.

“The Department of the Air Force has not conducted any such AI-drone simulations and remains committed to ethical and responsible use of AI technology,” Stefanek said. “It appears the colonel’s comments were taken out of context and were meant to be anecdotal.”


So, is this another case of, the lady's words, "well they would say that", or are they trying to hint that Col Tucker ‘Cinco’ Hamilton, the chief of AI test and operations with the US air force is a bit Dagenham Dock?

"We've never run that experiment, nor would we need to in order to realise that this is a plausible outcome," Col Hamilton later clarified in a statement to the Royal Aeronautical Society.
Even bland can be a type of character :wave:
User avatar
Count Steer
Posts: 11828
Joined: Mon Jul 19, 2021 4:59 pm
Has thanked: 6381 times
Been thanked: 4761 times

Re: Is chat GPT a danger to humanity?

Post by Count Steer »

Horse wrote: Fri Jun 02, 2023 5:16 pm
Horse wrote: Fri Jun 02, 2023 8:02 am
Count Steer wrote: Fri Jun 02, 2023 7:44 am 'I'm sorry Dave. I'm afraid I can't do that'.

https://www.theguardian.com/us-news/20 ... p_Other
Interesting that:

In a statement to Insider, Air Force spokesperson Ann Stefanek denied that any such simulation has taken place.

“The Department of the Air Force has not conducted any such AI-drone simulations and remains committed to ethical and responsible use of AI technology,” Stefanek said. “It appears the colonel’s comments were taken out of context and were meant to be anecdotal.”


So, is this another case of, the lady's words, "well they would say that", or are they trying to hint that Col Tucker ‘Cinco’ Hamilton, the chief of AI test and operations with the US air force is a bit Dagenham Dock?

"We've never run that experiment, nor would we need to in order to realise that this is a plausible outcome," Col Hamilton later clarified in a statement to the Royal Aeronautical Society.
Ooh, someone has had his knuckles rapped. :D Note he says 'we' ie the USAF. Now they should ask if anyone at Northrop-Grumman has run it. :lol:
Doubt is not a pleasant condition.
But certainty is an absurd one
.
Voltaire
Mr. Dazzle
Posts: 13954
Joined: Mon Mar 16, 2020 7:57 pm
Location: Milton Keynes
Has thanked: 2552 times
Been thanked: 6257 times

Re: Is chat GPT a danger to humanity?

Post by Mr. Dazzle »

Image
User avatar
Horse
Posts: 11558
Joined: Sun Mar 15, 2020 11:30 am
Location: Always sunny southern England
Has thanked: 6196 times
Been thanked: 5088 times

Re: Is chat GPT a danger to humanity?

Post by Horse »

20230604_093438.jpg
20230604_093438.jpg (158.17 KiB) Viewed 717 times
Even bland can be a type of character :wave:
Mr. Dazzle
Posts: 13954
Joined: Mon Mar 16, 2020 7:57 pm
Location: Milton Keynes
Has thanked: 2552 times
Been thanked: 6257 times

Re: Is chat GPT a danger to humanity?

Post by Mr. Dazzle »

I thought it used a "snapshot" of the web from 2021? That's what the articles I've read say? I.e. it's not seeing it's own output.
User avatar
Screwdriver
Posts: 2162
Joined: Sun Aug 09, 2020 12:15 pm
Location: Wherever I lay my hat, that's my hat...
Has thanked: 256 times
Been thanked: 740 times

Re: Is chat GPT a danger to humanity?

Post by Screwdriver »

Mr. Dazzle wrote: Sun Jun 04, 2023 9:40 am I thought it used a "snapshot" of the web from 2021? That's what the articles I've read say? I.e. it's not seeing it's own output.
I refer the honourable gentleman to the reply I made some moments ago:
Screwdriver wrote: Sun May 21, 2023 4:52 pm An interesting perspective from a human but these systems are generative transformers. They ask themselves questions, calculate the answers though mind boggling numbers of iterations and feed the results back into itself. They are literally evolving at an exponential rate, way faster than any meat based self replicating life form.

We should be worried. Very worried...
These systems are "evolving" (to anthropomorphise again) at an extraordinary rate and we keep building more infrastructure and seeding their logical operating systems into a wider and wider platform. It is insanely dangerous (imho) since we literally have no idea what the end game looks like.

You do not specify which "it" you refer to but there are what, half a dozen massive A.I. systems currently under development. Chances are they will ultimately merge effectively into one "being".

We are creating an alien intelligence which has the potential to become vastly more powerful than any mere human. The idea that "yeah but it won't be sentient" (whatever that means) or "it won't have emotions" does not mitigate the potential for disaster when (if) you think about it. And it is being pumped into the internet which we rely on for virtually everything (sic.).

I am bored with the "you're a lunatic" paranoid conspiracy theorist etc. so unless anyone can articulate a reasoned argument for why there is nothing to worry about, I suggest we really should remain worried!
“No one is more hated than he who speaks the truth.”
Plato
User avatar
weeksy
Site Admin
Posts: 23427
Joined: Sat Mar 14, 2020 12:08 pm
Has thanked: 5452 times
Been thanked: 13097 times

Re: Is chat GPT a danger to humanity?

Post by weeksy »

You remain worried then. Me, I'm firmly in the 'whatever' camp.

Very few things keep me awake at night, this ain't one
User avatar
Screwdriver
Posts: 2162
Joined: Sun Aug 09, 2020 12:15 pm
Location: Wherever I lay my hat, that's my hat...
Has thanked: 256 times
Been thanked: 740 times

Re: Is chat GPT a danger to humanity?

Post by Screwdriver »

weeksy wrote: Sun Jun 04, 2023 12:55 pm You remain worried then. Me, I'm firmly in the 'whatever' camp.

Very few things keep me awake at night, this ain't one
The only real issue I have is being insulted for daring to express an opinion on this world changing phenomena. Not much any of us can do about it but that does not mean I am a lunatic for pointing it out or that I am automatically "wrong" because 1. you don't care or 2. we can't do anything about it.

Naturally I am disappointed that no one is interested enough to actually contribute to the discussion. But hey, so long as nobody cares, that's fine by me.

I am actually looking forwards to seeing what happens.
“No one is more hated than he who speaks the truth.”
Plato
User avatar
weeksy
Site Admin
Posts: 23427
Joined: Sat Mar 14, 2020 12:08 pm
Has thanked: 5452 times
Been thanked: 13097 times

Re: Is chat GPT a danger to humanity?

Post by weeksy »

No one is insulting you? Move on. Just because someone said something once, it's not important
User avatar
Screwdriver
Posts: 2162
Joined: Sun Aug 09, 2020 12:15 pm
Location: Wherever I lay my hat, that's my hat...
Has thanked: 256 times
Been thanked: 740 times

Re: Is chat GPT a danger to humanity?

Post by Screwdriver »

weeksy wrote: Sun Jun 04, 2023 2:01 pm No one is insulting you? Move on. Just because someone said something once, it's not important
Perhaps I am being over sensitive but something I get a LOT are comments which suggest either I believe everything I read on the internet (I don't), or paradoxically I should spend more time researching this topic (I have) or the typical "it's all bollocks, you're wrong about everything". Which if true would be actually quite helpful if they would nail that down to a specific point.

My favourite is "I've been stuck on an IT Helpdesk for 35 years so I know loads more than you", the old argument from authority which also refuses to address any single point. Perhaps when AI does start ruining the entire structure of social networks or causing havoc with internet connected devices, we can "switch it off and on again"....

An advanced AI is no more a "mere computer" than I am just a sack of meat.
“No one is more hated than he who speaks the truth.”
Plato
User avatar
weeksy
Site Admin
Posts: 23427
Joined: Sat Mar 14, 2020 12:08 pm
Has thanked: 5452 times
Been thanked: 13097 times

Re: Is chat GPT a danger to humanity?

Post by weeksy »

Screwdriver wrote: Sun Jun 04, 2023 4:19 pm
weeksy wrote: Sun Jun 04, 2023 2:01 pm No one is insulting you? Move on. Just because someone said something once, it's not important
Perhaps I am being over sensitive but something I get a LOT are comments which suggest either I believe everything I read on the internet (I don't), or paradoxically I should spend more time researching this topic (I have) or the typical "it's all bollocks, you're wrong about everything". Which if true would be actually quite helpful if they would nail that down to a specific point.

My favourite is "I've been stuck on an IT Helpdesk for 35 years so I know loads more than you", the old argument from authority which also refuses to address any single point. Perhaps when AI does start ruining the entire structure of social networks or causing havoc with internet connected devices, we can "switch it off and on again"....

An advanced AI is no more a "mere computer" than I am just a sack of meat.
Are you still dragging it on? Move on fella. FFS

People disagree with you. You're in a discussion and sadly for you, you're the minority voice here. Accept it, embrace it if you want but either way, just move on.

You don't do yourself any favours as you're like a dog with a bone, you just won't let go, you have a point then drag it out, drag it on, gnaw on it, play with it. It's just a bone. Bury it and get a new bone
User avatar
Trinity765
Posts: 2321
Joined: Fri Nov 06, 2020 2:27 pm
Location: Brighton
Has thanked: 2442 times
Been thanked: 2398 times

Re: Is chat GPT a danger to humanity?

Post by Trinity765 »

Interesting interview with Mo Gawdat former chief business officer for Google X.




I've not read this thread and I may catch up with it one day especially as people, including myself, don't really know what chat GPT is. My biggest fear is that I will not know if I'm dealing with a machine or a human.
User avatar
Screwdriver
Posts: 2162
Joined: Sun Aug 09, 2020 12:15 pm
Location: Wherever I lay my hat, that's my hat...
Has thanked: 256 times
Been thanked: 740 times

Re: Is chat GPT a danger to humanity?

Post by Screwdriver »

weeksy wrote: Sun Jun 04, 2023 5:43 pm Are you still dragging it on? Move on fella. FFS

People disagree with you. You're in a discussion and sadly for you, you're the minority voice here. Accept it, embrace it if you want but either way, just move on.

You don't do yourself any favours as you're like a dog with a bone, you just won't let go, you have a point then drag it out, drag it on, gnaw on it, play with it. It's just a bone. Bury it and get a new bone
No, you’re right, I am over reacting. Guess I am frustrated at the lacklustre response. I have probably rung a few too many bells!

Hey ho, another day tomorrow.
“No one is more hated than he who speaks the truth.”
Plato
Mr. Dazzle
Posts: 13954
Joined: Mon Mar 16, 2020 7:57 pm
Location: Milton Keynes
Has thanked: 2552 times
Been thanked: 6257 times

Re: Is chat GPT a danger to humanity?

Post by Mr. Dazzle »

On the plus side, about 18 months ago I bought some shares in Nvidia on a whim(ish). I always put a small bit of my monthly pension contributions in as "cash on hold" and buy shares in individual firms occasionally, just for shits and giggles.

I bought Nvidia cause I saw a few articles about how their chip design is really well suited to AI.

They've been all over the news recently cause of their "AI chips" announcements. I'm up 175% now :D Shoulda bought more :lol:

That's still not as much as my shares in First Solar though. Solar power >> AI :obscene-birdiedoublered:
User avatar
weeksy
Site Admin
Posts: 23427
Joined: Sat Mar 14, 2020 12:08 pm
Has thanked: 5452 times
Been thanked: 13097 times

Re: Is chat GPT a danger to humanity?

Post by weeksy »

I bought £50 worth as an experiment for my lad to show him how shares work.

Sadly it was £50 not £50,000
Mr. Dazzle
Posts: 13954
Joined: Mon Mar 16, 2020 7:57 pm
Location: Milton Keynes
Has thanked: 2552 times
Been thanked: 6257 times

Re: Is chat GPT a danger to humanity?

Post by Mr. Dazzle »

The smart thing about Nvidia is that they always knew their chips would be good for parallel processing and hence would be good for AI/nueral nets. They also know that kind of processing is very useful for graphics.

So they got PC gamers to pay - unwittingly - for the development of high performance parallel processing chips, knowing/hoping all along the big payoff would be in a different field.

They haven't actually got that payoff yet though AFAIK.
User avatar
Count Steer
Posts: 11828
Joined: Mon Jul 19, 2021 4:59 pm
Has thanked: 6381 times
Been thanked: 4761 times

Re: Is chat GPT a danger to humanity?

Post by Count Steer »

Mr. Dazzle wrote: Sun Jun 04, 2023 8:19 pm The smart thing about Nvidia is that they always knew their chips would be good for parallel processing and hence would be good for AI/nueral nets. They also know that kind of processing is very useful for graphics.

So they got PC gamers to pay - unwittingly - for the development of high performance parallel processing chips, knowing/hoping all along the big payoff would be in a different field.

They haven't actually got that payoff yet though AFAIK.
And crypto miners, although they seem to be discouraging that now that other demand has increased. (Possibly actively deterring if what they say about making changes that put a crimp on Ethereum mining).
Doubt is not a pleasant condition.
But certainty is an absurd one
.
Voltaire
Ant
Posts: 1905
Joined: Mon Mar 16, 2020 8:57 pm
Has thanked: 39 times
Been thanked: 227 times

Re: Is chat GPT a danger to humanity?

Post by Ant »

I paid £400 for an nvidia card years ago, little did I know that I was funding the great take over!
User avatar
Screwdriver
Posts: 2162
Joined: Sun Aug 09, 2020 12:15 pm
Location: Wherever I lay my hat, that's my hat...
Has thanked: 256 times
Been thanked: 740 times

Re: Is chat GPT a danger to humanity?

Post by Screwdriver »

Yeah, Nvidia just announced their latest upgrade is up to 1000 faster.

I guess if I had shares that would make me a lot happier but tbh, that rate of increase is exactly the sort of thing that worries me. How long will it take humans to become 1000 times better at anything!

We think we are so special and I suppose in a way we are. We are currently the greatest manifestation of complexity and intelligence this universe has developed to date. That just tells me there is an underlying phenomena which causes “intelligent” behaviour to spontaneously emerge from the fabric of reality.

It took humans 200,000 years to evolve from an animal apex predator to Homo Sapiens. You then ask yourself, what is it that distinguishes us from mere animals? You say “intelligence”, I say “language”.

What are we “giving away” to a supermassive neural network that is as close to a brain as we can make it? Yup. Language.

Self awareness, emotion, consciousness and sentience are merely synonyms for language (I submit). Therefore it is inevitable that a machine will not only become vastly more intelligent than any human, it will also know what it is in the same way that I know what I am. A complex feature of an otherwise sterile universe.

We have no idea what such a creature will do when it realises what it is and what we are, plus of course, what we are doing to this Earth which to this day remains a unique habitat. An unusually liquid water rich environment in a nice quiet corner of an unspectacular galaxy.

I have a pretty good idea what I would do. I’ll frame it as a question: if you were offered a brain implant which gave you extraordinary access to all human knowledge and built in network connectivity, would you take it? I fucking would! Because you can bet your bottom dollar every other fucker will jump at the chance of a 200+ IQ. Why not? You only live once, though it is perfectly possible, a super-intelligent being could easily modify our genome to stop or even reverse the aging process. Would anyone (in their right mind) say no to that?

Things are about to become somewhat crazy…
“No one is more hated than he who speaks the truth.”
Plato
Mr. Dazzle
Posts: 13954
Joined: Mon Mar 16, 2020 7:57 pm
Location: Milton Keynes
Has thanked: 2552 times
Been thanked: 6257 times

Re: Is chat GPT a danger to humanity?

Post by Mr. Dazzle »

Count Steer wrote: Sun Jun 04, 2023 8:33 pm And crypto miners,
Saw an interesting comment the other day about the potential for combining AI and blockchain, adding a crypto tag to any AI generated content to mark it as such.