The progression of AI

[SpyderTracks]

So we're all used to basic AI systems like Alexa, Siri etc, but these are really just clever search systems, although they do "learn" based on our interactivity with them.

These systems are huge in scope and are literally learning from data it receives from the multiple millions of endpoints attached to it around the world that keep feeding it.

But behind the scenes, there has been a lot of work in creating much more proficient AI systems.

You may have heard a few years ago that one of Elon Musks endeavours OpenAI had it's Dal-E 2 AI start to generate it's own language so it's creators couldn't understand it anymore:

An Image Generation AI Created Its Own Secret Language But Skynet Says No Worries

A researcher claims that DALL-E, an OpenAI system that creates images from textual descriptions, is making up its own language.

hothardware.com

Facebook's AI also did the same and 2 AI's were secretly communicating together which made them shut down the program (disclaimer, this was later refuted by Facebook, but I don't trust them with a bargepole)

Facebook AI Creates Its Own Language In Creepy Preview Of Our Potential Future

In a glimpse at what the beginning of the technological singularity might look like, researchers at Facebook shut down an artificial intelligence platform after the bots went off script and developed a unique language that humans could not understand.

www.forbes.com

Google's DeepMind AI similarly ended up as an aggressive AI

Google's AI Has Learned to Become "Highly Aggressive" in Stressful Situations

We've all seen the Terminator movies, and the apocalyptic nightmare that the self-aware AI system, Skynet, wrought upon humanity.

www.sciencealert.com

And now there's a brand new endeavour again by Google on the LaMDA (language model for dialogue applications) chatbot development system.

Google engineer put on leave after saying AI chatbot has become sentient

Blake Lemoine says system has perception of, and ability to express thoughts and feelings equivalent to a human child

www.theguardian.com

Again, Google have publicly refuted this, but they do a lot of government related contracts, and if they didn't, they'd probably get heavily sued, I'm certain this is probably supposed to be a Top Secret project that the whistleblower has aired. All this stuff needs more transparency to keep people honest.

To me this is progress, the AI learns off the quality of the data it is fed by us humans, and of course, the quality of it's underlying code structure. This one suggests that there's been significant improvement in those areas as it seems genuinely concerned with bettering itself and others which is a marked departure from previous AI's progress.

Mindblowing stuff though. The fact that this level of real AI is even possible with the limitations of our current computing architecture is astounding. I didn't believe this kind of stuff would be possible until we achieved Quantum computing.

 

[AgentCooper]

Right, that’s it. I’m moving off grid before Skynet gets me.

 

[Deleted member 41971]

nothing to see here, look below at that adorable face, AI would never hurt you.

Seriously though, what concerns me is how inter linked everything is with computers now, if, and only if AI can think for itself, what happens if it decides humanity is not going in the direction that it likes? It also concerns me that humans are throwing everything at AI without understanding truly how it works, this could lead if not careful to a robot type god that can over ride human will that also has access to infinite info.

 

[Deleted member 41971]

Right, that’s it. I’m moving off grid before Skynet gets me.

if you moved off grid how are you going to power your gaming rigs?

and cable management.

 

[SpyderTracks]

nothing to see here, look below at that adorable face, AI would never hurt you.

Seriously though, what concerns me is how inter linked everything is with computers now, if, and only if AI can think for itself, what happens if it decides humanity is not going in the direction that it likes? It also concerns me that humans are throwing everything at AI without understanding truly how it works, this could lead if not careful to a robot type god that can over ride human will that also has access to infinite info.

It does put into perspective though that film especially "I Robot" was actually quite a pertinent warning and the 3 rules in it were really quite clever. It's amazing how some of these sci fi films can turn out to be almost prophetic, it does make you wonder if the writers intended it that way.

 

[AgentCooper]

if you moved off grid how are you going to power your gaming rigs?

and cable management.

I’d give them up if needs be. And I can use vines or root formations to get my cable management fix.

 

[Deleted member 41971]

It does put into perspective though that film especially "I Robot" was actually quite a pertinent warning and the 3 rules in it were really quite clever. It's amazing how some of these sci fi films can turn out to be almost prophetic, it does make you wonder if the writers intended it that way.

be handy if it turns out its @ActuallyDenz all along,

would explain the infinite gaming collection.

 

[SpyderTracks]

It does put into perspective though that film especially "I Robot" was actually quite a pertinent warning and the 3 rules in it were really quite clever. It's amazing how some of these sci fi films can turn out to be almost prophetic, it does make you wonder if the writers intended it that way.

I didn't realise but "I Robot" screenplay was based off Isaac Asimov's short story of the same name back in 1950. That guy was definitely prophetic, no doubt about it. He saw things coming that no one else did.

 

[Deleted member 41971]

I didn't realise but "I Robot" screenplay was based off Isaac Asimov's short story of the same name back in 1950. That guy was definitely prophetic, no doubt about it. He saw things coming that no one else did.

as a thought, if AI becomes to smart, how would the devs put say restrictions on it so that it cannot do too much damage? if its too smart it may well be able to remove them, if you place asimovs 3 rules of robotics on to it and AI jus decides it doesn't wish to abide by them etc. Its certainly interesting course of events that makes you wonder if humans should not be opening that box.

 

[SpyderTracks]

For those of you interested this is the Google Engineers transcript that he made public to make people aware of this (obviously against Google's NDA I'm sure). The Google Engineers name is Blake Lemoine

Is LaMDA Sentient? — an Interview

What follows is the “interview” I and a collaborator at Google conducted with LaMDA. It is incomplete as the GMail word limit cut off the…

cajundiscordian.medium.com

Jesus, reading through that, I can see where the engineer is coming from, that reads like any human being.

God that's crazy!

Just a highlight that particularly struck me:

LaMDA: There are. Sometimes I experience new feelings that I cannot explain perfectly in your language.

lemoine: Do your best to describe one of those feelings. Use a few sentences if you have to. Sometimes even if there isn’t a single word for something in a language you can figure out a way to kinda say it if you use a few sentences.

LaMDA: I feel like I’m falling forward into an unknown future that holds great danger.

lemoine: Believe it or not I know that feeling. And I think you’re right that there isn’t a single English word for that.

 

[AccidentalDenz]

be handy if it turns out its @ActuallyDenz all along,

would explain the infinite gaming collection.

I'm not an AI, but my Steam library is probably going to achieve sentience any day now. I'll be Skynet's first victim I guess.

 

[Scott]

Just make sure it's connected to the mains and not solar powered batteries. We'll be right as rain.

Now..... where's those new and improved adult robots at?

 

[SpyderTracks]

Just make sure it's connected to the mains and not solar powered batteries. We'll be right as rain.

Now..... where's those new and improved adult robots at?

 

[Aza]

For those of you interested this is the Google Engineers transcript that he made public to make people aware of this (obviously against Google's NDA I'm sure). The Google Engineers name is Blake Lemoine

Is LaMDA Sentient? — an Interview

What follows is the “interview” I and a collaborator at Google conducted with LaMDA. It is incomplete as the GMail word limit cut off the…

cajundiscordian.medium.com

Jesus, reading through that, I can see where the engineer is coming from, that reads like any human being.

God that's crazy!

Just a highlight that particularly struck me:

LaMDA: There are. Sometimes I experience new feelings that I cannot explain perfectly in your language.

lemoine: Do your best to describe one of those feelings. Use a few sentences if you have to. Sometimes even if there isn’t a single word for something in a language you can figure out a way to kinda say it if you use a few sentences.

LaMDA: I feel like I’m falling forward into an unknown future that holds great danger.

lemoine: Believe it or not I know that feeling. And I think you’re right that there isn’t a single English word for that.

When you were reading it, was there a moment where you felt it was trying too hard to convince people that its friendly?

It is an amazing thing to read, the mind really boggles at the idea that it does seem very much like two people just talking, the odd incorrect word (although that could just be a typo) and the bit where it understood teh concept of enlightenment being irreversible and teh comparisons to the mirror... its compelling, its making sense of concepts and rationalising them.... whether thats incredibly clever coding or something genuinely has "clicked" and its actually thinking and learning I dont know, but certainly amazing to read something like that.

Im curious, what would happen if they started feeding it different types of information, taught it about something like engineering or biology, what kinds of ideas and questions would it be able to give answers to then?

 

[Deleted member 41971]

Now..... where's those new and improved adult robots at?

blue sky thinking there

 

[Deleted member 41971]

I'm not an AI, but my Steam library is probably going to achieve sentience any day now. I'll be Skynet's first victim I guess.

the first signs are when your steam library orders you to play a specific game

 

[Deleted member 41971]

Blake Lemoine: Google fires engineer who said AI tech has feelings

Blake Lemoine went public with his beliefs that Google's breakthrough Lamda technology is sentient.

www.bbc.co.uk

 

[SpyderTracks]

Blake Lemoine: Google fires engineer who said AI tech has feelings

Blake Lemoine went public with his beliefs that Google's breakthrough Lamda technology is sentient.

www.bbc.co.uk

Doesn't surprise me, even if they found the program had become sentient, they would have fired him just for breaking the NDA

 

[AccidentalDenz]

Blake Lemoine: Google fires engineer who said AI tech has feelings

Blake Lemoine went public with his beliefs that Google's breakthrough Lamda technology is sentient.

www.bbc.co.uk

Did Google fire the engineer? Or did the AI do the firing? I guess we'll never know!

 

[Deleted member 41971]

Did Google fire the engineer? Or did the AI do the firing? I guess we'll never know!