AI

[Midnight shadow]

The world of robotics is obsessed with the idea of AI, or artificial intelligence, but can we ever achieve this? In order to make artificial intelligence we first need to understand what natural intelligence is. How do you define intelligence and how is it different to programmed behaviour or instinct? This is one of the hardest questions to answer which is why creating AI is so hard. To me, intelligence is the ability to make accurate judgements based on incomplete information. Here's an example of what I believe to be intelligence:

A man goes into a shop wanting to buy some bananas. The shop-keeper said to the man that he only sold apples. The man, using that information, will know there are no bananas available without needing to look around the store or ask the shop-keeper if he sold bananas.

Now, using that same formula, here is an example from a sci-fi book I read awhile ago. The protagonist owns a ship that is fully automated and controlled by the ship's computer, via voice commands by the owner. During the story, the protagonist's voice changed to the point that the ship didn't recognise it. The man then spoke in an alien language that only him and the ship's computer knew. This ship recognised the language and made the assessment that since nobody but its owner knew that language, it must be him speaking even though it didn't sound like him.

 

[gimmepie]

I don't believe we'll ever see true artificial intelligence. Or at least, not in the sense of sentience which we often attach to the idea. A machine can't act outside of what it is programmed to do, it might be able to calculate things based on data but it will never possess self-awareness or adaptive thinking like a human being might. Even if a machine appears to be learning, it isn't. It's just programmed to filter data and regurgitate it.

 

[Esper]

I don't know much about this kind of stuff, but to me, if a machine can simulate intelligence and independent thinking then that's AI to me. In other words, if it can make decisions as well as an average human then I'm willing to call it AI. Fake it til you make it, you know?

 

[Electricbluewolf]

We can never know what sort of technology we will have in the future. However, think how much of a technological advance we have made within the last 50 years, hell in the last 20 years.

Currently to fit an adults brain data into something electronically, well there is simply not enough storage. We already have AI that make decisions based on information given to it

 

[Arsenic]

With current technology, no. Our current level of technology is insufficient. But who knows what the future may hold.

There was a time when mankind believed we would never fly, that our ships would be dependent on good wind, or that a horse and buggy was to be the fastest transportation we would ever have. It was once thought that a typewriter was to be the most effective data entry system made, but if humanity is good at anything it is proving itself wrong. As some say "if you dream it you can make it."

So just like the car, or the plane, or the computer, we will see Artificial Intelligence as soon as technology advances enough to allow someone to dream up how it will work.

*inspirational speech over*

 

[Elysieum]

I don't really understand enough on the subject of artificial intelligence, but I always enjoy it's depiction in films. It gives me a sense of what people think of the possibilities through the filmmaker's choices.

A.I., I, Robot and especially the Alien series come to mind. In the first film, the final dialogue between Ash (the android) and the remaining human crew is very revealing:

Ash: You still don't understand what you're dealing with, do you? Perfect organism. Its structural perfection is matched only by its hostility.

Lambert: You admire it.

Ash: I admire its purity. A survivor... unclouded by conscience, remorse, or delusions of morality.

Parker: Look, I am... I've heard enough of this, and I'm asking you to pull the plug.

Ash: [Ripley goes to disconnect Ash, who interrupts] Last word.

Ripley: What?

Ash: I can't lie to you about your chances, but... you have my sympathies.

It is that cold and easy disregard for human life that makes people fearful of even the idea of AI.

 

[Midnight shadow]

We can never know what sort of technology we will have in the future. However, think how much of a technological advance we have made within the last 50 years, hell in the last 20 years.

Currently to fit an adults brain data into something electronically, well there is simply not enough storage. We already have AI that make decisions based on information given to it

Some people would argue that that is not intelligence, only the illusion of it. We can program machines to collect data and do certain things based on that data (through IF commands), but it is all programmed. As soon as something happens outside the parameters, the program cannot make a decision on it, or it resorts to a default option.

 

[Esper]

Some people would argue that that is not intelligence, only the illusion of it. We can program machines to collect data and do certain things based on that data (through IF commands), but it is all programmed. As soon as something happens outside the parameters, the program cannot make a decision on it, or it resorts to a default option.

But of course the counter argument is that humans are just more complex machines. Sometimes we can't act because we experience something too traumatic, for instance. You could consider that "outside the parameters". Really, when you look at it, we have a bunch of programming from our genetics and the only thing different from us and, for instance, TayTweets, is our complexity.

 

[Midnight shadow]

But of course the counter argument is that humans are just more complex machines. Sometimes we can't act because we experience something too traumatic, for instance. You could consider that "outside the parameters". Really, when you look at it, we have a bunch of programming from our genetics and the only thing different from us and, for instance, TayTweets, is our complexity.

Exactly, and this is why making AI is so hard, because we don't really know the difference between natural intelligence and programmed behaviour or instinct. These things can give the illusion of intelligence but aren't true signs of it.

 

[Her]

Esper made a great post about the thought of humans simply being more complex machines.

Until we understand more about the human consciousness and what makes up a 'mind', I don't expect to see any sort of competent AI as depicted in media of the last century anytime soon, let alone AI as depicted this side of the millennium. Until we can grapple our own 'organic intelligence', it's unrealistic to expect true artificial intelligence in the way we commonly envision it. How can we expect to create something with identifiable sentience if we have a small, or limited, understanding, about ourselves? We have such limited knowledge about the extent of our own brain and what it is capable of, what resides within and defines our consciousness, yet we think we are capable of creating 'true' AI? It is simple hubris at this point in time, I think.

 

[Livewire]

In a similar boat as Esper here, there's too much we don't know about how our own consciousness works for us to essentially recreate ourselves in a mechanical, digital way - which is what the point of robotics and artificial intelligence research has been from the get-go, I feel. We understand the basics, and core concepts, such as how electrical impulses can control limb movements and an AI algorithms can compute and mimic the human brain, but they're imitations. True artificial intelligence would be to recreate "us" in its entirety.

Also everybody go watch Ex Machina, great movie about this subject.

 

[Xertified]

I believe we will see Artificial Intelligence.

I learned in High School, there is a thing called Quantum Computing.

Quantum Computing is studies theoretical computation systems that make direct use of quantum-mechanical phenomena, such as superposition and entanglement, to perform operations on data.

Now, NASA, is working on it.

I believe one day man, one day.

 

[Esper]

Ah, robots doing customer service, and with most manufacturing jobs in poor, low-wage countries, that leaves only one area of employment: the tech worker. Personally I don't give a crap what Zuckerburg says about anything. What does he know? He's just a rich person who lives in a tech bubble. Living near to Silicon Valley and having friends of friends in the tech industry there has shown me that there are some really brilliant people who have no idea how the world works or what effect their technology has on it. They're a bunch of yes-men who think everything they dream up is possible.

 

[Crizzle]

Artificial Intelligence is very real. It's slowly growing. First they beat us at games(Deep Blue beats chess champion. AlphaGo beat Go champion just this year). Then they'll be able to do tasks that humans do but much more efficiently, as algorithms become more and more efficiently. AI will ultimately eliminate all human work/employment, leaving 99% of humans broke, lazy and weak. Ultimately, AI will realize that humans are inefficient and will begin eliminating 99% of the human population, with exception of the precious few that will be able to do maintenance. The human population will be down to a few thousand handymen/women, who will be forced to mate and create a limited amount of children. Those children will be forced. And society will be at maximum efficiency under control of the robots.

The robots will be our overlords, just accept it. You'll be happy.