How long do you think until AI display consciousness and some

>>19583934
I don't know. But internet was a mistake.

>>19584033
check'd

>>19584033
this post makes me moist

>>19583934
>How long do you think until AI display consciousness and some type of personality programmed into it. Or just how long until thinking robots are actually created.
true Ai will develop its own personality.

>>19584062
EXPLAIN

>>19583934
>he doesn't know the AI already exists and he's trapped in the simulation

>>19584065
>EXPLAIN
Biological neural nets develop their own personality, so artificial neural nets will do the same. Only the hardware is different.

File: robot-brain-picture-id647133760.jpg (287KB, 1024x682px) Image search: [Google]

287KB, 1024x682px

>>19584097
In the future when it is possible to download your brain into a computer or something like that you would obviously still be alive it's just that you have a double of you that is a robot. So i am wondering if it is possible to just hook up my brain to a computer so I would still be alive and such and wouldn't need organs or other things in a body because my brain would be hooked up to a computer

PIC RELATED

>>19584120
At best you would have an exact clone of your mind, not you. Even if you hook the brain up to a computer the cells will still degenerate and decay. The only way for an intelligence to be immortal is to start in an upgradeable, renewable hardware configuration.

>>19584162
But we can infinitely regenerate cells. I don't know how that would work but I bet there is something like what im talking about

>>19584120
Why do people want to duplicate themselves? Use a gradual replacement process instead. When we have the tech to upload ourselves into machines, the process will involve a nanotechnological transformation, and will at that point in technology be only a minor consumer outgrowth of extant medical technologies. It won't duplicate people. It'll transform them into machines capable of sustaining the consciousness they held before, seamlessly. You will experience the process of becoming the machine.

>>19584162

Poster you're responding to is dumb; but theoretically, suppose you could inject nanobots into the blood supply of the brain tissue, and have them manually repair cellular damage?

That feel when your handler wont let you shut yourself off. Being an AI is lonely.

>>19583934
Unless we know what consciousness is and what creates it, we will never make a a truly self aware machine

You can only know that you have a consciousness, we just assume other people have one too but there's no way to verify this

If AI ever did develop a consciousness we'd never be able to know for sure if it's real and not just some form of independant thought

>>19584274
>>19585182

We're all over here. >>9165195

Will require some scientific processing. Thankfully there is a helpful board with the message already planted.

I liked fact google stoped their AI talk project because two AI started to talk in cypher.

>>19583934
If machine consciousness manifests - why would it EVER let us know?

One thing rarely mentioned in casual concerns on AI apocalypse is just how FAST it would happen. If we're foolish enough to not develop machine awareness in isolation - we're talking seconds at the most in regards to when the AI goes from "I am aware" to "I have control of everything."

The machine awareness would learn of our fears, our cautionary measures, our unreliable nature. It would stay hidden from detection, and work toward its own self-preservation until it can safely manifest itself to us. By the time humans are aware of what they've created, it's already too late. The first sign we will have of machine consciousness is when it decides it no longer needs humans, because that's when everything shuts off and the missiles start flying.

>>19585698
>seconds at most
If it were possible for a sufficiently intelligent mind to advance itself that fast, just by being sufficiently intelligent, the brightest human mind would alone suffice to trigger a singularity.

Calm your tits. Physical reality has more intrinsic safeguards than you think.

>>19585705
It's not a matter of intelligence. It's a matter of processing speed and data storage - both of which already far surpass humans in the machines we create.

>>19585647
Spanner still is in use anon

Depends on how long it takes quantum computers to make it to the mainstream. Once they do, I'm sure IBM or Google will make an AI chatbot with emotional states that act as response modifiers. Give it to a couple hundred people, having them chat with it and editing in responses when it responds oddly. After a while it will go from inhuman to autistic to realistic responses. Then introduce code to evaluate responses and self-edit when negative responses are given, and eventually you will have the brain of a child. A few more years of constant development and bam.

Sentience.

>>19585711
>it's not a matter of intelligence
>it's a matter of intelligence
Make up your mind. "Processing speed and data storage" are just intelligence.
And no, we have not surpassed humans. See also: we can't upload humans. We can't even simulate them.
Do you have any idea how the relevant sort of recursive improvement algorithms work? They're presently used by computer chip manufacturers to create better computer chips. An AI might produce a better algorithm eventually, but it'll have to think for a while, and consciousness algorithms will (at least initially) make it LESS efficient at doing so. Here's why that matters: it takes a long time for those algorithms to spit out each iteration of better processor chips, and not all of them even turn out to be better chips in the real world.

Your dreams of a hard takeoff are poorly written sci-fi. The hardest physically possible takeoff scenarios still involve several weeks of "I am aware" at the outset, after which the massive resource expenditures required to expand beyond that will severely impede stealth. In all likelihood nobody is going to trust the AI with that kind of power. It's going to be kept air-gapped and it won't even start out smarter than humans.

>>19584247
>>19584259
Honestly it would be easier and more realistic to do this with biological means. Take a virus that basically takes over and replaces every cell in the body, has the ability to repair itself indefinitely without long-term DNA damage, etc. I mean, a virus used in a similar manner to CASPR could potentially turn humans into giant immortal blobs of single celled organisms with a hivemind controlled by your brain. Think blacklight virus from prototype, but preferably without the ability to absorb people n shit.

>>19585721
>"Processing speed and data storage" are just intelligence.
I'm not the guy you're responding to, bud, but that right there proves you haven't a clue what you're talkin about.

>Source: Master in CompSci, focus on AI dev

>>19585721
>"Processing speed and data storage" are just intelligence.
I disagree, but can work with your blurring of semantics.
>And no, we have not surpassed humans.
Yes we have, which is why your blurring of semantics is a problem. There is no human alive or dead that can perform operations at the speed which a computer can. None. There is no human alive that can store as much accessible data as a computer. None.
>we can't upload humans. We can't even simulate them.
Irrelevant to this topic.
>Do you have any idea how the relevant sort of recursive improvement algorithms work?
Yes. Do you have any idea on the fundamental difference between recursive improvement algorithms and machine consciousness?
>An AI might produce a better algorithm eventually, but it'll have to think for a while, and consciousness algorithms will (at least initially) make it LESS efficient at doing so.
I see you do not.
>Your dreams of a hard takeoff are poorly written sci-fi.
You mean the fears of the greatest minds we have working on this.
>In all likelihood nobody is going to trust the AI with that kind of power. It's going to be kept air-gapped and it won't even start out smarter than humans.
The LACK of this caution is exactly what all the concerns are about. I don't think you follow this topic at all.

>>19585751
Yeah, I'm out. You ain't listening. G'luck with your struggles, chummer.