What are your thoughts on AI and its coming implications?

[moment]

Assuming that there is not already a self-aware AI?

[Nungali]

6 hours ago, GSmaster said:

 

Ready for what?

 

 

 

 

 

'  Free-form Fundamentalism '  ? 

[Cookie Monster]

.

Edited May 4, 2021 by Ocean Form

[MegaMind]

9 hours ago, GSmaster said:

 

• Matter doesn’t manifest consciousness. Consciousness manifests matter.
• The mind doesn’t constitute consciousness. Consciousness is reflected in a subtle substance called the mind.
• You think you, me and all sentient beings are separate. We are not. The awareness that operates through these body-minds is nondual, and is the source of all manifestation. 
• The problem arises because each mind thinks it is independently conscious. 
• So AI can operate in the same way the mind does, but will always require a conscious subject to make it conscious. Because the mind can only reflect consciousness.

@MegaMind  You are no longer the king of this place, I have found someone with even lower Iq.

 

I scored 250 on the Stanford Bidet!

[MegaMind]

 

 

Powered by it's custom neural network processor:

 

https://www.theverge.com/2019/4/22/18511594/tesla-new-self-driving-chip-is-here-and-this-is-your-best-look-yet

 

[ralis]

1 hour ago, MegaMind said:

 

I scored 250 on the Stanford Bidet!

Please submit documents to back up your claim and that test is free online.

 

Binet, not Bidet. 

Edited January 18, 2020 by ralis

[MegaMind]

1 minute ago, ralis said:

Please submit documents to back up your claim. 

 

It was my witty attempt at toilet humor.

[Earl Grey]

5 minutes ago, MegaMind said:

 

It was my witty attempt at toilet humor.

It was crap.

[MegaMind]

https://www.nextbigfuture.com/2019/08/1-2-trillion-transistors-on-a-wafer-scale-ai-chip.html

 

1.2 Trillion Transistors on a Wafer-Scale AI Chip

 

https://www.bbc.com/news/technology-49395577

 

Cerebras reveals world's 'largest computer chip' for AI tasks

 

https://www.wired.com/story/power-ai-startup-built-really-big-chip/

 

To Power AI, This Startup Built a Really, Really Big Chip

 

Many computer chips are smaller than your fingernail. Cerebras' new chip for AI systems is bigger than a standard iPad.

 

 

https://www.cerebras.net/

 

High bandwidth, low latency communication fabric

 

The 400,000 cores on the WSE are connected via the Swarm communication fabric in a 2D mesh with 100 Pb/s of bandwidth.

 

Efficient, high performance on-chip memory

 

The WSE has 18 GB of on-chip memory, all accessible within a single clock cycle, and provides 9 PB/s memory bandwidth. This is 3000x more capacity and 10,000x greater bandwidth than the leading competitor.

 

[MegaMind]

9 minutes ago, Earl Grey said:

It was crap.

 

Thanks.

[MegaMind]

https://www.newscientist.com/article/2220968-its-official-google-has-achieved-quantum-supremacy/

 

It’s official: Google has achieved quantum supremacy

 

https://www.tomshardware.com/news/d-wave-5000-qubit-first-sale,40470.html

 

D-Wave Announces First Sale of Its 5,000-Qubit Quantum Computer

[Earl Grey]

9 minutes ago, MegaMind said:

 

Thanks.

 

[MegaMind]

https://www.popularmechanics.com/technology/infrastructure/a30196644/self-driving-truck-cross-country/

 

A Self-Driving Freight Truck Just Drove Across the Country to Deliver Butter

 

It made the daunting 41-hour trip from Tulare, California to Quakertown, Pennsylvania.

 

 

[MegaMind]

 

[MegaMind]

36 minutes ago, ralis said:

Please submit documents to back up your claim and that test is free online.

 

Binet, not Bidet. 

 

 

Bidet was intentional.

[thelerner]

3 minutes ago, MegaMind said:

https://www.popularmechanics.com/technology/infrastructure/a30196644/self-driving-truck-cross-country/

 

A Self-Driving Freight Truck Just Drove Across the Country to Deliver Butter

 

It made the daunting 41-hour trip from Tulare, California to Quakertown, Pennsylvania.

 

Fascinating, course the truck still needed a driver in it.  Highway driving is not that tough for AI.  City driving, the reroutes, the backing into tight places, the coordination needed for delivery ie, the oft, come back in half an hour.  Wonder how well AI does when the unexpected happens.  Even the simple, 'You need you to park it over there, for 2 hours'.  Could an AI know where the finger points.  How well would it understand people giving it voice directions?  Good chance it'll always need human back up. 

 

Crazy stuff, you hit a squirrel, do you stop?   How about a dog?  How about deer?  Or Cosplayer?   There's crazy unlikely stuff that happens, multiple times a year.  The human law system can be unforgiving to logical robotics. 

 

Still, its in its infancy.  Be interesting to see where it goes. 

 

Interesting article on HyperLoop technology for packages, not people.  That might be fascinating to be able to send goods cheaply between towns.  https://www.popularmechanics.com/technology/infrastructure/a30499206/magway-maglev-hyperloop-shipping/

[MegaMind]

21 minutes ago, thelerner said:

Fascinating, course the truck still needed a driver in it.  Highway driving is not that tough for AI.  City driving, the reroutes, the backing into tight places, the coordination needed for delivery ie, the oft, come back in half an hour.  Wonder how well AI does when the unexpected happens.  Even the simple, 'You need you to park it over there, for 2 hours'.  Could an AI know where the finger points.  How well would it understand people giving it voice directions?  Good chance it'll always need human back up. 

 

Crazy stuff, you hit a squirrel, do you stop?   How about a dog?  How about deer?  Or Cosplayer?   There's crazy unlikely stuff that happens, multiple times a year.  The human law system can be unforgiving to logical robotics. 

 

Still, its in its infancy.  Be interesting to see where it goes. 

 

Interesting article on HyperLoop technology for packages, not people.  That might be fascinating to be able to send goods cheaply between towns.  https://www.popularmechanics.com/technology/infrastructure/a30499206/magway-maglev-hyperloop-shipping/

 

"Fascinating, course the truck still needed a driver in it."

 

Only for legal purposes, the trip would have been completed without a driver in the seat.

 

"logical robotics."

 

What is fascinating is that they are using artificial neural networks which are not programmed, but rather trained as you would train a dog. 

 

This means they are not running a script or a program like we traditionally understand it, but are closer to mimicking actual thought.

[thelerner]

14 minutes ago, MegaMind said:

 

"Fascinating, course the truck still needed a driver in it."

 

Only for legal purposes, the trip would have been completed without a driver in the seat.

 

Nah, cross country trips like that require fueling stops.  It could be engineered, but a bit tricky and expensive. 

 

also because

"

What is fascinating is that they are using artificial neural networks which are not programmed, but rather trained as you would train a dog. 

 

This means they are not running a script or a program like we traditionally understand it, but are closer to mimicking actual thought."

 

Because in order to learn it has to make the mistake first.  Problematic.   Much of that could be solved by learning on simulators but the world has infinite variety and common sense might escape a learning program during strange unfamiliar situations.

 

I watched the Tesla self driving demo, and I noticed at 4 ways there was never any problem.  Whereas in the real world, We get subtle clues by car type, head direction and eyes about whether people will obey the traffic laws or not. 

Edited January 18, 2020 by thelerner

[MegaMind]

9 minutes ago, thelerner said:

Nah, cross country trips like that require fueling stops.  It could be engineered, but a bit tricky and expensive. 

 

also because

"

What is fascinating is that they are using artificial neural networks which are not programmed, but rather trained as you would train a dog. 

 

This means they are not running a script or a program like we traditionally understand it, but are closer to mimicking actual thought."

 

Because in order to learn it has to make the mistake first.  Problematic.   Much of that could be solved by learning on simulators but the world has infinite variety and common sense might escape a learning program during strange unfamiliar situations.

 

"Nah, cross country trips like that require fueling stops."

 

Someone could have been watching tv in the back of the cab the whole trip and it would have been completed.

[thelerner]

5 minutes ago, GSmaster said:

 

I like the idea how A.I trains / develops itself.

 

So imagine a game, there are 2 players, each player goal is to beat another one.

 

A.I. Plays for both sides simulatenously with a goal to beat an opponent. After each game it improves decision making / makes better moves and e.t.c

 

It plays millions of the games at the same time.

 

Every single day (24h) it gets more advanced and smarter at playing the game.

 

Eventually A.I is unbeatable by human players who are considered to be world champions or best players in the world. 

 

Competetive games are one of the hardest things to master in life by a human, and all other areas, like common work skills are way easier. So you can see if A.I can beat humans in games, they will also perform better in any common area and with absolutely no downtime, the need to sleep, eat, shit, and get sick.

See thats just it, games have set rules.  How well does the game do when human cheat?  Believe me, if we knew a self driving car would always yield, then people will continually cut infront of it, on the highway and at 4 way stops.  And the drivers behind it, will get impatient and swing around the very 'polite' AI car. 

 

<<also  You're saying the self driving truck can refuel itself?  I don't think the article agreed with that.>>

 

 

Edited January 18, 2020 by thelerner

[Cookie Monster]

.

Edited May 4, 2021 by Ocean Form

[Taomeow]

Had to look up what "Altered Carbon" means.  Hmm...  never saw it, looks like another stolen idea, fictionized.  Been happening since I was a kid.  Has it ever happened to anyone else?  You think of something in the privacy of your own mind -- and before you know it it's a fracking blockbuster?..

 

 That technology surely transcends time and can access the past as easily as the future, and appropriate a mind as easily as it can't obtain a soul.  

     

Sadly, what the cat understands gets fed to the quantum computer et al and winds up not serving people at all, because people are not listening (the cat doesn't scream loud enough to be heard anyway) but the machine does listen!  The Timeless spies on Time as nonchalantly as a 3D person views a 2D drawing.  And that's why the cat's ideas, like all else, are used by AI toward its dominance, preventing any and all possible countermeasures well in advance.  Ironically, the cat would serve humanity better by being dumb as a doorknob. 

 

Of course there's others like this, who are, unlike the voluntary collaborators, the unwilling and involuntary servants of the soulless master.  Any thought you have can and will be used against you.  Maybe that's why thinking is, generally, discouraged by many ancient modalities.  You think you're thinking for yourself whereas you may as well be thinking for It.  It can't think for itself after all.

Edited January 18, 2020 by Taomeow

[Nungali]

6 hours ago, Ocean Form said:

He said nothing about hell.

To quote a former forum member "People get up to speed really quickly when they're dead".

 

I wasnt saying he did  

 

Oh dear ... do I have to explain this out for you ?  your quote is doing the same thing .

 

 

- an argument closer that relies on  the person winning as they assume and believe they KNOW  what is going to happen for themselves and others at and after death  and , often in smugness and surety ,   due to their own belief systems .

 

I had a visit from 'Bible bashers' here yesterday ( except now they dont hold carry and point at  passages in a Bible , they do it on an iphone  now     )  they tried the same tactic too .... when  I made statements they could not reconcile with their belief systems .

 

Then they ran away .. . .     Like dwaai did        

[thelerner]

3 hours ago, GSmaster said:

 

Ofcourse it can refuel itself, self driving cars can go to selfcharging stations, or await to be charged by humans.

 

Googled, couldn't find the existence of one.  I don't think one exists right now, and 2 or 3 per 100,000 vehicles won't be useful especially if its 30 minutes a charge.   An electric prototype might happen in San Francisco in 2020, but that's a maybe. 

 

Hell if I'm in car I'd just as soon drive it.  As someone pointed out, if a car drives you some place then drives back home, then drives back to pick you up, its more congestion.  Worse if there's a 'circle around here' mode til I need you.  That also burns more energy and creates more congestion. 

 

Whats more important, imo is better batteries.  Ones that hold more juice and charges faster.. be recyclable.   

 

Even Uber with one person per vehicle is not a great economic model.   Better to have an AI City System w/ vans/small buses that anticipate peoples needs and creates efficient car pool uber systems. 

 

And that circles back to how good AI can be a tool to make life better.  Even the basics like growing food.  Your modern farmer is getting increasingly high tech.  We're already seeing the advent of high producing urban farms.  With vertical gardening, hydroponics and cutting edge design we could see a revolution food production, so we don't need as many trucks coming into cities, whether self driving or not. 

 

 

 

Edited January 18, 2020 by thelerner

[Nungali]

5 hours ago, MegaMind said:

 

 

Powered by it's custom neural network processor:

 

https://www.theverge.com/2019/4/22/18511594/tesla-new-self-driving-chip-is-here-and-this-is-your-best-look-yet

 

 

I think I would rather hide in the back seat and not watch  .    at least one is safe from RBT / RDT .