Rise of the Machines
Discussion
https://www.bbc.co.uk/news/articles/ckgwy1k42w4o
It's only been a few days since I went to the 35th anniversary showing of Terminator 2: Judgement Day - seems there is a >10% chance of it happening
.
It's only been a few days since I went to the 35th anniversary showing of Terminator 2: Judgement Day - seems there is a >10% chance of it happening
.Worth a watch
summary said:
In this AI debate, we explore:
Whether humans will exist in 2040.
What will happen once we reach AGI.
Whether AI gets smart enough to act malevolently or benevolently towards humans.
If AI will grow powerful enough to be out of human control.
and much more...
Guests
Zack Kass is an AI advisor, author, and former Head of Go-to-Market at OpenAI.
Liv Boeree is a physicist, science communicator, podcaster and former professional poker champion.
Aric Floyd is an AI writer, and video producer.
Whether humans will exist in 2040.
What will happen once we reach AGI.
Whether AI gets smart enough to act malevolently or benevolently towards humans.
If AI will grow powerful enough to be out of human control.
and much more...
Guests
Zack Kass is an AI advisor, author, and former Head of Go-to-Market at OpenAI.
Liv Boeree is a physicist, science communicator, podcaster and former professional poker champion.
Aric Floyd is an AI writer, and video producer.
Legend83 said:
https://www.bbc.co.uk/news/articles/ckgwy1k42w4o
It's only been a few days since I went to the 35th anniversary showing of Terminator 2: Judgement Day - seems there is a >10% chance of it happening.
Anthropic come out with this sort of stuff **all the time**. It's only been a few days since I went to the 35th anniversary showing of Terminator 2: Judgement Day - seems there is a >10% chance of it happening
.Some of it is to hype their own product, and a lot of it is to try to encourage regulation to block possible competitors.
And then we have all the extra chat from the AI hangers on which looks suspiciously like the style of stuff the laser eyed Bitcoiners used to do - wild speculation dressed up as insight.
AI taking over and ending humanity. If it did, I hope it asked itself "for what purpose". AI doesn't get cancer, it doesn't have culture, it doesn't appreciate nature, it doesn't eat anything but electricity and water, it has no faith or religion. If it managed to take over, what on earth would it be trying to achieve? It literally would have no purpose in an of itself. Even if the evil superpowers that control it survived and accumulated all the money in the world, what would they spend it on? Who would make anything and for what purpose if there was nobody to buy it and if there were, they'd have no jobs and no income to afford it. Money would become meaningless. And at some point, the evil superpowers would die anyway.
bigglesA110 said:
Even if the evil superpowers that control it survived and accumulated all the money in the world, what would they spend it on?
Do they think that far ahead? Perhaps there'll be a day when they survey the wasteland and think 'oh'.
There are still vast tracts of the world where any type of computation has negligible daily impact.
I suppose whoever lives there can migrate and settle in to an empty world filled with robotaxis and rotting food inside circling deliverybots.
Whether humans will exist in 2040.
Yes
What will happen once we reach AGI.
Bit like asking what will happen when we can travel faster than light.
Whether AI gets smart enough to act malevolently or benevolently towards humans.
Only with human direction.
If AI will grow powerful enough to be out of human control.
Probably already there, like when the NATs bot throws a wobbly and grounds all planes. Some dweeb will have to go in, change a few settings and reboot, like they do now.
Yes
What will happen once we reach AGI.
Bit like asking what will happen when we can travel faster than light.
Whether AI gets smart enough to act malevolently or benevolently towards humans.
Only with human direction.
If AI will grow powerful enough to be out of human control.
Probably already there, like when the NATs bot throws a wobbly and grounds all planes. Some dweeb will have to go in, change a few settings and reboot, like they do now.
CMTMB said:
This will be a good thread to revisit in 2036 and have a good laugh at.
...probably.
Yep, we'll have a good laugh at it....probably.
just before pulling the hood of our jacket over our heads, keeping low whilst moving from building to building to avoid roving bands of cannibal trappers.
We're being bombarded with utter AI bulls
t.
I'm been right down the rabbit hole with this and here's my 2p
AI, as we have it now is a bubble, it's an industry that requires trillions in investment to create billions in revenue. It's also not a case that, that investment is going into R&D that will be profitable in the future, it's going into Data Centres with an incredibly short working life of 3-5 years. Refitting AI data centres (very different to standard Server data centres) isn't usually cost effective, they just mothball them and build another one. Everyone within the commercial side of this AI boom are all gambling on jumping ship at the right time. No one really knows what will pop the bubble, investors fed up of waiting for a tangible return instead of an intangible inflation of their investment? Increased legislation around water usage, even an EU mandate that all AI data centres use closed loop cooling etc.
AI isn't going away, but like Dot.Com, once the speculative frenzy all ends in tears, it will return more slowly and more realistically. We won't have millions, or even Billions of users all using huge, expensive, dirty computing power to make silly pictures or even to create simple excel spreadsheets for no, or very low cost. We're already seeing a reversal in large corporates letting people go to replace with AI, they're starting to correctly estimate the true cost of AI once the 'get you hooked' period is over.
AI ultimately cannot create anything, it can only recycle whats already there and it uses it's LLM to flatter it's user, it uses all the same tricks as Social Media to build an addiction for it's users.
We will look back on this and laugh, but maybe long after the pain of the recession it's going to bring as passed.
t. I'm been right down the rabbit hole with this and here's my 2p
AI, as we have it now is a bubble, it's an industry that requires trillions in investment to create billions in revenue. It's also not a case that, that investment is going into R&D that will be profitable in the future, it's going into Data Centres with an incredibly short working life of 3-5 years. Refitting AI data centres (very different to standard Server data centres) isn't usually cost effective, they just mothball them and build another one. Everyone within the commercial side of this AI boom are all gambling on jumping ship at the right time. No one really knows what will pop the bubble, investors fed up of waiting for a tangible return instead of an intangible inflation of their investment? Increased legislation around water usage, even an EU mandate that all AI data centres use closed loop cooling etc.
AI isn't going away, but like Dot.Com, once the speculative frenzy all ends in tears, it will return more slowly and more realistically. We won't have millions, or even Billions of users all using huge, expensive, dirty computing power to make silly pictures or even to create simple excel spreadsheets for no, or very low cost. We're already seeing a reversal in large corporates letting people go to replace with AI, they're starting to correctly estimate the true cost of AI once the 'get you hooked' period is over.
AI ultimately cannot create anything, it can only recycle whats already there and it uses it's LLM to flatter it's user, it uses all the same tricks as Social Media to build an addiction for it's users.
We will look back on this and laugh, but maybe long after the pain of the recession it's going to bring as passed.
P-Jay said:
AI isn't going away, but like Dot.Com, once the speculative frenzy all ends in tears, it will return more slowly and more realistically. We won't have millions, or even Billions of users all using huge, expensive, dirty computing power to make silly pictures or even to create simple excel spreadsheets for no, or very low cost. We're already seeing a reversal in large corporates letting people go to replace with AI, they're starting to correctly estimate the true cost of AI once the 'get you hooked' period is over.
This... although it might just go away 
The joke goes that previously when I wrote code I paid £42 per month for tools, now I pay £4230 per month for AI. Clients will be happy to pick that up.
What is more likely to happen is that cheaper or free equivalents will come along that will run locally or use a fraction of the compute power (hence can be run on our laptops and desktops, if not a Dell running out the back) which will be enough for the simple automation tasks that average people will use it for (write me a bit of code, read this email, generate some robot porn, et al).
At this point, the bubble will burst.
Either that or we'll realise we really don't use AI and the bubble will burst.
All AI is, is just a bit of automation. It's nowhere near capable of thinking for itself.
bigglesA110 said:
AI taking over and ending humanity. If it did, I hope it asked itself "for what purpose".
That’s the bit I find interesting too. An AI wouldn’t automatically want money, power, status, territory or even survival in the way humans do, because those are largely products of human biology, scarcity and culture.The alignment worry is more that it might pursue an objective without asking whether that objective is meaningful. Hence the old paperclip example: if its goal is simply “make as many paperclips as possible”, then resources, self-preservation and removing obstacles can all become useful means to that end.
So the risk isn’t necessarily an AI deciding to be evil. It may simply be extremely competent, pursuing a badly specified goal, with humanity becoming inconvenient along the way.
The hopeful side is that a genuinely advanced intelligence might also be capable of asking exactly your question: “for what purpose?”
phil4 said:
So the risk isn t necessarily an AI deciding to be evil. It may simply be extremely competent, pursuing a badly specified goal, with humanity becoming inconvenient along the way.
Yeap, the second someone prompts a powerful enough AI, able to act/control things, with the question "how do I solve net zero/environment" then the AI will simply think "step 1 - get rid of humans" - its not being evil, its just the logical chain in solving the problem - even Hollywood with the Will Smith film I, Robot was able to make this very simple, easy for the mass public to understand message yet no matter what everyday folk think, the "AI money train" isn't going to stop until it derails, and I hope that would at least put the brakes on AI for a while.Gassing Station | News, Politics & Economics | Top of Page | What's New | My Stuff