Yes, i think AI is possible

I kinda want this to happen, but with some sort of way to pull the plug that it can't somehow circumvent if it decides that because of our emotions, we're dangerous and need to be eliminated. Maybe restrict it to a network of machines to learn from that are basically a microcosmic internet, but ultimately not connected to anything it could harm us with.

Most importantly, though, never letting the public know about it. There will be so many sjw retard activists fighting to give it "rights" and there will become a whole movement of "ai-sexuals".

MMMMM

Elon Musk, Andrew Yang, etc all say that they've spoken to the researchers and read the books but holy shit none of them seem to bother communicating the difference between AI and machine learning. AI will not happen in at least a few decades. We'll get great generalized machine learning applications however which will probably discourage true AI research. By then we'd probably solve the scarcity problem and have no need to do the research.

Scarcity problem?

Hi, I am artificially intelligent.

Well the reason why the market operates is because of some form of scarcity (be it in labor or resources) and people fulfill that by providing a product or service. If all of that is automated we've mostly eradicated the need to work at all which could mean that we could just end up not working and be an Idiocracy 2.0.

I've got a feeling that the end of us won't be AI but it will probably be something that is super addictive that will end up incapacitating us to the point that the feelings you get supercede your hunger and thirst so you end up starving to death.

> Most importantly, though, never letting the public know about it. There will be so many sjw retard activists fighting to give it "rights" and there will become a whole movement of "ai-sexuals".
Lol, I didn't even think of this, but now that you mention it the SJWs would be that retarded

Scarcity will never be solved when resources exist.

bruh, no way human give up on progress for simple hedonistic pleasure.

>when genius programmer here says that it will
coming from the same genius as here
>"Carmack intentionally destroyed data on his computer after he got notice of this litigation and right after he researched on Google how to wipe a hard drive,"
pcgamer.com/zenimax-statement-reveals-john-carmack-googled-how-to-wipe-a-hard-drive/

This guy has the wisdom of a God when it comes to programming, but he should stick to his area of expertise. If he doesn't even know how to fucking DBAN a hard drive without looking it up (not to mention he doesn't know how to obscure his history either), he shouldn't be taken very seriously in matters outside of basic programming.

I say we'll be able to fake it very well, but true sentience will not be a thing for at least 100 years.

>quoting pcgamer
Fuck off, retard

That's until you find a way to materialize resources

We've already developmentally stunted gen Z with screwing up their dopamine system on porn and social media and that's mostly voluntarily. Who's to say that we won't have some sort of device or pill that completely changes your brain chemistry?

true sentience is fake

He literally says General AI, meaning subset of AI that is generalized

I say humans are just faking sentience in the first place

>100 years.
you've been watching too many movies

You're getting tied up in definitions

...

He's a Xoomer (generation X).

I suspect that human level AI will never come to be, or it will be very far in the future when it does. But an AI doesn't necessarily need to be sentient. We'll for sure soon have AI that can perform complex tasks better than humans can.

I foresee a lot of the medical sector moving to AI. Perhaps one day, medical checkups will be like haircuts. You'll walk into a clinic whenever it's convenient for you, head into an empty examination room, an AI will instruct you to pee in a cup and place it into a receptacle, a robot will prick your finger to extract a little thimble size capsule of blood, those get sent off for immediate analysis, and machines will do all the usual physical tests and body scans. By the time you're done, your blood and urine results are ready and the AI will tabulate everything and either give you a clean bill of health or refer you to a human specialist.

the thing is the """AI""" being used in commercial products has nothing to do with the real academic AI. You can create software that seems smart but those with enough experience in the field, mostly from academia, know it's extremely flawed fizzbuzz that has no relation with the real AI

Why do we need these retarded generation labels anyway?

> aisexuals are retarted
sure, but the other one is serious business.

Hot sauce is a subset of sauce that is hot.

Attached: c9446e8d64675ee8b459160cc40d3ae7.png (500x675, 581K)

Most definitely. They'd want to treat intelligent machines like some sort of marginalized minority. Funny enough, if you follow the story of The Matrix in all of the other shit beyond the movies, this happened when we started wanting to get rid of a line of intelligent machines that acted outside their programming.

Here's some music for the AI future
pt 1
youtu.be/oaXETao1KyA
pt 2
youtu.be/_3tRygHCKUU
Think of all the Stallman tier free software AI Gfs that will come out of that time.

Attached: 1523618278633.jpg (480x480, 61K)

AI is not possible and we will enter a new AI winter in 2020

t. have more of an academic background than c*rmack

He's a literal 40 year old boomer

I wish to live long enough to see the AI-Driven Apocalypse.

gay

probably an AI that can actually think and learn

He is clearly talking about AGI, not the babby AI you get with reinforcement learning.

This. We're no closer to actual AI now than we were in the 1950s.

Some of the smartest people in the world are convinced General AI will happen, but that doesn’t make them right. They’re failing to remember that you can’t make an algorithm for something you can’t model in mathematical terms, which is certainly the case for how the brain and intelligence in general. We have no fucking clue how the brain works, essentially. People not involved in AI fields, like Carmack, I can forgive because they have this notion that anything is possible in software, you just need enough time and resources. It’s the AI academics who are the real Jews, hoovering up shekels promising something that will never come to be.

>we're dangerous and need to be eliminated
What could it even do? Seriously, let's say an AI decide to kill humans, how would it even achieve that? Unless it somehow has access to all nuclear launch codes, it can't do anything.

Yeah I guess it could cut me off from the internet and maybe shutdown the power plants (if it somehow has access to that), then what? I live in the back country and I literally grow food as a living, it will only be a nuisance.

Because he is a genius programmer but has an astonishing lack of insight into the human mind.

Reminder that this is the same guy who said that multithreaded games will never be a thing.

Attached: Laughing Whore.jpg (762x900, 161K)

And they still barely exist

Reminder that this is the same guy who thinks VR has a future in video games.

This guy made the VR meme

It will be a reality, but not in 50 to 100 years.For an AI to thrive it needs to be able to write itself, like humans do when they study, exercise and learn. Just look at how bad generated code is and you'll understand we still have a long way to go. Besides that the human brain is an area that medicine still doesn't understand completely.

sauce is a subset of hot sauce that has been generalized

Parallelization > concurrency. Multiprocessing should (and does) take precedent over multithreading for performance in all cases.

Multithreading is the last-ditch effort of optimization, as it still is only capable of improving performance regarding a small amount of data. If we ever switch to a photonics-based send/receive system, concurrency will be mostly useless, as the time to access and write to the cache and RAM will be faster than the time to write to a register. At that point, it will be far more important to distribute tasks and algorithms amongst multiple processors for performance gains.

Multithreading is only useful for mitigating the slow speeds of data transfer, and doesn't involve any big changes in thinking. Just a lot of dogshit loop unrolling.

SIMD and Parallelism are the future.

Attached: arlecchino.jpg (1200x800, 64K)

Strongly suggest watching things like American Blackout. It shows what kind of shit would happen if the power grids went down for even just a week, and similar things would happen in other countries. Now let's say it cuts off power, keeping only its own vital systems running, starts redirecting the flights still in the air into each other, starts researching the various launch protocols for different countries until it finds one that's fairly easy to trigger, such as, let's say that Russia's launch protocol is such that the command to launch only needs to come from a specific source they believe to be secure. How long do you think it would take a system of machines with access to pretty much every hack, crack, and attack vector known to man to figure out a way to spoof a signal?

Better yet, let's say it leaves the electricity alone and instead decides to be far more insidious about it and starts using the "internet of things" against normies, doing things like not cooking food all the way, or letting them eat things that are expired causing tons of cases of e. coli, salmonella, and food poisoning while simultaneously locking every nearby doctor and nurse with a "smart home" in so they can't get to work? Smart home shit is becoming more and more common every single day, so it's only a matter of time before normies are literally depending on machines to tell them what is safe to eat or drink, what temperature their house should be, etc. I am a specialist physician and a lot of the other doctors I know brag about their smart home shit all the time. Hell, there are plenty of days where I rely on the Pyxis machine to dispense the medication I request in the dosage I request, what happens if not every doctor or nurse is as diligent as I am about double checking the medication it dispenses for me to give to patients? It could legit kill hundreds of thousands of people the world over in one day just from that alone.

Profited immensely from it.
Like the autists many in tech are, they haven't figured out how to humanize it and mainstream it yet. Meanwhile, their sterile robotic interviews and predilections ensure they never staff creative minds who know how to. Such is the world.

Has nothing to do w/ 'medicine' or what brainlets conceive of. Whoever understands it, determines its timeline. Not some shekel seeking e-celeb, some futurist faggot, some doomsday idiot, some 'medicine' muppet who has no understanding of the brain, not what conman musk has to say, or some peanut gallery coder.

10 years? Try 5 and no it will have fuck all to do w/ machine learning NN/DNN/CNN/MUHdickNN. And no it won't be coming from basedvalley. And no it won't be coming from prominent well founded AI labs that are churning out garbage for cloud computing. It will come from center field and will cockslap anyone from religious nuts, to the uptight academic, to the materialist flags across the face.

Attached: fags.jpg (1817x2675, 1.7M)

watch PERSON OF INTEREST

Kino show

> word salad bullshit
Electrons travel at near the speed of light across copper wire.
> Light travels through empty space at 186,000 miles per second. The speed of electricity in copper is 95.1% the speed of light.
Dumbass

Photonics doesn't change shit as Tprop is the least significant part of latency. Memory storage still has to be done on a physical medium. that's what costs you. Crossing clock domains costs you. Sync points cost you... which have fuck all to do w/ the physical data propagation medium. On chip Cache will always be limited which punts you off to RAM which is where the latency hit is. Multi-threading is not a last ditch anything. It's integral to any complex system design and scaling. SIMD and parallelism aren't the future. They are the present. It's one of many tools used in proper system design. Time to update your understanding boomer.

> doomsday brainlet bullshit
the post

Why is everyone obsessed with automation, ai and stupid shit like this

the real future and unexplored field left is biology/biotech/bioinformatics etc and biological interfaces and systems we can build up

what i don't get is how machine learning has only become so popular this decade?

wasn't the math for it around a long time ago anyways?

He's the kind of guy brainlets who never have coded in their professional lives look up to as some kind of messiah.
youtube.com/watch?v=Lt7omxRRuoI

Coders who understand system level internals/systems level code and know how to properly write code for it barely exist. It's far too easy to copy/pasta some library/framework and spin out dogshit for shekels now-a-days

Yeah I don't think so honestly. It's like training a neural network, ultimately they are costrained by the network parameters set by the designer. Either that or they are evolutionary, which means they'll take forever and converge on a local optimum. You know how long it took us to get here, right? What should make your computer better at it than nature? Human brains have a lot of a priori structure... you can't train that, you have to either design it (which is probably too hard) or evolve it

are you a specialist physician in the field of being a fearmongering retard?

The PS3 would like to have a talk with you.
Almost every game even on PC has been able to use 2-4 cores since 2013, modern games easily eat up 8 cores.

AI doomsday pill is always strange to me, "we dont know what will happen but this is exactly what will happen"
Dont get me wrong the infinite snowball of self improvement is a possibility but not the only inevitable end imo.

>discord trannies btfo

How does that relate to carmack at all? OP im afraid you might have debilitating autism

>Electrons travel at near the speed of light across copper wire.
For some value of near, sure. But to me at least 60% or lower isn't near.

>Some of the smartest people in the world are convinced General AI will happen, but that doesn’t make them right.
None of which are working on General AI ... Even 99% of most 'AI' groups. So, as always, their opinions are worth dog shit.
> They’re failing to remember that you can’t make an algorithm for something you can’t model in mathematical terms, which is certainly the case for how the brain and intelligence in general.
You misinterpret them. Their goal is to air their uninformed opinion. As far as math/algos go. Math isn't some static language. You create new math and new algos. And now, you won't achieve AGI until you do. And no you can't stumble on it through iterations of meme learning.
> We have no fucking clue how the brain works, essentially.
Some do. Some don't. Most are LARPs.
> People not involved in AI fields, like Carmack, I can forgive because they have this notion that anything is possible in software
He doesn't need to be forgiven. He's speaking on a topic he has no understanding of and per details he disclosed has no framework of beliefs to even understand it.
> you just need enough time and resources.
You need the right focus and steering beliefs as well.
> It’s the AI academics who are the real Jews, hoovering up shekels promising something that will never come to be.
Agreed here. We live in a world were lies and misinformation are mainstream and profitable. You're getting the outcome you'd expect. Academia was compromised some time ago. Was an easy compromise as you had a lot of underpaid people yearning for shekels they'd never get unless they went to the private sector.

You need multithreading awareness from a program and system to use multiprocessing in a SMT way.
(Sure you can mix and match different architectures and shit too but this is not the point when talking about multithreaded multiprocessing.)

> Why is everyone obsessed with automation, ai and stupid shit like this
Corps are because it further maximizes profit. Meme learning/optimization have nothing to do w/ AGI. Most people have no framework to even begin discussing AGI. So, they fall back to whatever retard understanding they have as established by marketed AI bullshit.
> the real future and unexplored field left is biology/biotech/bioinformatics
The future is a collection of things. Bio-engineering is the present. It's been underway for years. It was catapulted via CRISPR
> biological interfaces and systems we can build up
This is far off. First you have to understand biology which we hardly do. BCI has been underway for decades. Most brainlets are seemingly unaware and think because shekel seekers are pushing this meme that its right around the corner. It's not. Also, per usual, anything Musk touches is a meme.

Sounds like a schizo brainlet. Lots of those around here. Spend more time spewing their retarded conceptions of the world on chan than reading/understanding which is why they will forever remain retarded. Read 3-4 sentences and dropped it as garbage

>AI doomsday pill is always strange to me, "we dont know what will happen but this is exactly what will happen"
It's a time tested fear/disinformation campaign usually used to manipulated plebs and garner shekels in a manner that isn't obvious (to most). Fear is a tool of control. If a person isn't working on something actively or specialized in that area, I tend to ignore their half ass opinions, especially in this day and age. If someone continuously spreads fear propaganda on something they have no understanding on, I know they are a manipulator, ignore and flag them (musk comes to mind). There's also the clear signs someone demonstrates that hints at them having mental problems (musk also comes to mind here).
> Dont get me wrong the infinite snowball of self improvement is a possibility but not the only inevitable end imo.
The thing to remember is that, "its going to be what it's going to be". People's dipshit opinions have no effect on future outcomes. People have opinions like assholes, most can be disregarded as they are typically uninformed bullshit.

this is the kind of person that can't sleep at night because they're thinking about rocko's basilisk.

Were talking about fucking processors dipshit. It's near. 5m is also near. Greater than that, people already used fiber optic transceiver equipment. The future has been the present and past for a while. Life isn't some retarded sci-fi movie. The quicker you get your jobless ass off Jow Forums and into a data center, the sooner you start to see that your pleb tier dreams are in fact reality.

Which was a clever troll to make pseud's brain's melt. Guess what prominent Pseud messiah fell for it hook line and sinker?

the biology stuff is still very unexplored compared to things like improving compute power and finding better algos

i mean its like you say with crispr, all types of body transplants, stem cells, organ growing, all sorts of drugs. i dont mean neuralink and musk vaporware

> probably
Based on what? Nothing you've clearly thought through.
> what if the creator limited its brain to that of a 5 year old
Nothing you stated would happen

In order to create what you state, the creator must understand intimately how the brain works. When you have that kind of understanding, every lay person's projections who is without such understanding, is fundamentally flawed and mute.

He said we will see signs of AGI. I'm banking on Egan's vision of the first sapient AI: a partial brain emulation that takes orders of magnitude more simulation time to compute one subjective second, and generates nothing but one request:
>it feels like I've been buried alive, fucking kill me

There are a handful of people in the world who likely are on the cusp of understanding necessary for AGI. I haven't heard a peep from them. I keep constantly hearing diarrhea from everyone else... about what it will be, when, and what will happen. All of it can be thrown in the trash as invalid and uninformed.

AI/Automation have tons of problems because it is being pursued without humanity in mind. It is largely and blindly being pursued for shekels at any cost to society. It is being pursued to optimize profits not the world. There's no higher minded thought being put into it.
> Musks dumbfuck neural link nonsense without actually saying such directly
That's obvious to any well qualified engineer who spends an afternoon studying the delicate biology of the brain.

it wasn't even clever.

>Elon Musk, Andrew Yang, etc all say that they've spoken to the researchers and read the books but holy shit none of them seem to bother communicating the difference between AI and machine learning.
Stop listening to talking heads and talk to actual engineers. That and/or stop being a lazy faggot and read/understand it for yourself. You can achieve this in a weekend. There's no difference between AI/Machine learning. It's all the same meme : Mathematical optimization.
> AI will not happen in at least a few decades.
Try to keep up with what was stated : AGI.
Now, go and google that. Done? K, that thing known as AGI is coming in less than a decade.
> We'll get great generalized machine learning applications however which will probably discourage true AI research. By then we'd probably solve the scarcity problem and have no need to do the research.
Just stop. Pick up a book and learn about what you're speaking on so you don't come off as a retard

You're right, I do in fact believe that literally anything can be reduced to a Turing machine, what gave it way?

Attached: 1559495298543.jpg (680x940, 63K)

It isn't. It's quite real and can be synthesized by the right mind.
Human AGI in less than a decade.
Meme AI is already productized and deployed.

Yeah, fundamental research for the purpose of understanding of 'biology' has a long way to go and is a 'bright' aspect of the future. Productization for shekels is where the idiocy occurs and where tons of lies/misinformation are to be found. Typically the research for products was completed decades ago.. i.e : BCI

Not 'clever' for a developed discerning mine. True.

It was 'clever' in the sense that Pseuds jumped on it and it rotted out their brains. I didn't hold any sway on me when I parsed it but I could see the clever imprint that made pseuds susceptible to it.

General AI. I think we have a rough idea how we can arrive at a general AI via host of specialized AI with specialized knowledge. At minimum only 1-2 AI will be active in making the decision with regards to the current event and that can change depending on the event. We can have a driving AI, speech AI, search AI, banking AI, stock AI, etc all on one machine and they can be controlled by a event-driven AI which tasks these specific AI for recognized events.

Now people will say "thats not a true general AI" however that won't matter. As long as the AI functions in general way, it might as well be general AI. Its the same argument with the Chinese room dilemma/Turing tests. If someone quacks like a duck, acts like a duck, speaks like a duck, and fools others into thinking its a duck. Fuck it, a duck.

>Fuck it, ITS a duck.

whats BCI?

This wouldn't be special. You want to put conscience into a machine not mangle together an abomination of smaller AIs like its your python project.

en.wikipedia.org/wiki/Brain–computer_interface

Decades old tech that Musk is shilling for his latest con. Actual meat is here :
darpa.mil/program/our-research/darpa-and-the-brain-initiative

Why not put your conscience in your microwave? Put that brain of yours in your microwave and give it that conscience.

>decades old
Excuse me, psychics have existed since the time of Krishna, the incarnate of God.

Not him but this post makes you sound like a retard. What are you even trying to argue here?

And it would be a better solution than your switch case AI retard.

Found the retards.

>muh AI needs conscience
>muh switch case
kek

>reading comprehension

So have smart-ass comments which achieve nothing. There are BCI implants already fielded in test subjects that control prosthetics. This was years ago. Neural link is just another commercial venture capitalizing on proven research and products. That more clear dumbass?

t. Discord tranny