[Kappers]

Good day all,

I'm a new member from London England, and am likely to be younger than most at only 17. I currently have little theoretical knowledge of artificial intelligence as I am yet to attend university, however I aim to attend a high achieving institute to study artificial intelligence in computer science and learn it there! I do however have quite an expansive philosophical knowledge.

I work with Java, and have become pretty competent however I don't have a firm grasp on advanced techniques, and advanced graphical techniques such as using the 3D libraries. I aim however to extend my Java knowledge through learning more about AI, the subject I adore! A few unrelated projects of mine include a set of user driven physics simulations, including a spring engine and kinetic theory of matter educational program. I also am planning and fleshing out concepts for an emotionally driven game (something I believe vital for progress in the game industry, as it takes dynamic content to the next level), I obviously cannot go as far as producing it in this current state  .

At my college I'm also working on what we call an "Extended Project", a qualification whereby we write a university standard dissertation. My project title is, "Can we justify our attempts to create automatons with emotional intelligence?". It is a heavy philosophical discussion which evaluates a lot of literature, I have therefore studied a lot of the work of early AI pioneers and done a lot of further reading which has spurred me on to join this forum, whereby I aim to maintain my enthusiasm for the area of science.

Although we have not seen true AI, and I don't believe we will due to paradoxes proposed by Moravec and Samuels, I believe to succeed in driving short term artificial intelligence we must invent in the employment of underlying human intelligent devices such as emotion, or invest in the potential of in vitro culturing of brain tissue. However I won't bore you any more!

Chat soon,
Tom 

[infurl]

http://en.wikipedia.org/wiki/Moravec's_paradox

Moravec's paradox is based on a fallacy, and even if it wasn't, it's not a paradox. It is merely an observation of difficulty, not a proof of impossibility, and the worst that it implies is that we have to go back to the drawing board. If true, it wouldn't be the first time that it took human ingenuity several tries to get something right.

He's saying that solutions to the simplest tasks have been evolving for the longest time and will therefore be the most difficult to reverse engineer. That assumes that solutions to simple tasks are never good enough and they have to keep evolving, which isn't true. Evolution is continuous because the environment changes. Once evolution finds a solution, that feature stops evolving until some change in the problem which caused it occurs. The change in the environment may be so great that the evolution of that feature becomes irrelevant altogether, something which has happened time and time again for as long as life has existed.

The other big mistake that he makes is in thinking that things like logic and reasoning are always simple. The process may always be simple. For example, logical resolution which is the basis of theorem proving is very simple. However the complexity of problems varies. Some problems can be solved very quickly and some problems could not be solved before the end of the universe because of the number of steps or permutations involved. Many of these "simple" problems such as the well known Travelling Salesman problem are what is called NP-Hard (look it up). Biology and evolution have not found solutions to these kinds of problem either -- nature has to do exactly the same thing that we have to program our computers to do to find the answers: guess.

By the way, welcome to the forum, I am sure that you (and we) will have an interesting time.

[DaveMorton]

Hi, Tom! Welcome aboard!

I have to disagree with your supposition that we'll never see "true AI". In fact, I believe this is something that will be achieved, likely within my lifetime, though likely not within the next ten years or so. With the breakthrough of IBM's Watson, and explosion of advances over the past few years that have increased computing power and speed, I believe that we're slowly building toward a new era of discovery that will net some real, positive results toward the advancement of what I like to term as "Synthetic Intelligence". I doubt, however, that we'll achieve "machine sentience" within my lifetime. That, I fear, is still a long way off.

[Kappers]

Moravec's paradox is based on a fallacy, and even if it wasn't, it's not a paradox. It is merely an observation of difficulty, not a proof of impossibility, and the worst that it implies is that we have to go back to the drawing board. If true, it wouldn't be the first time that it took human ingenuity several tries to get something right.

I do agree with what you're saying to an extent, however even observations of difficulty have in the past been the most prominent barriers to scientific progression. Unlike barriers such as an ever-increasing need for powerful processing devices, which I believe is comparably trivial, understanding ourselves fully may prove to be as difficult as what we find around our universe. I don't have time to go into it now, but I definitely want to explore this topic in the future, I have quite a lot of reasoning to reinforce what Moravec proposed, although obviously it's not perfectly succinct!

I have to disagree with your supposition that we'll never see "true AI".

Well to say we'll never see "true AI" depends on what you consider to be true, it's inevitable there will be conflict in finding a definitive explanation of what is "true AI" as there is already such ambiguity surrounding how we define AI itself. In my mind, without having to consider the rationale behind behavior and thinking, for our purposes as humans, it is merely supposed recreation of human intelligence. And for us to do that, we'll need a full understanding of how us as humans function, for example we'll need deep understanding of our cognition. Understanding this will mean learning, to then fully recreate this gained intelligence in automatons we'll need to further our knowledge even more, and the process repeats (and lets not add evolution to the mix). Although I won't deny we as humans may eventually see " true AI ", I don't think I'll ever live to see it. I do appreciate this argument is flawed, but the proof is in the pudding!

[Art]

Welcome aboard, Tom. Nice to have you with us and as always, new ideas and discussions
help the rest to not become stale or complacent in our own thoughts of how things should be.

We may not see these things in our lifetimes and perhaps we're not quite ready for it, but we
might see the cusp or leading edge of where this technology is headed. Only time will tell.

Have you read much into Ray Kurzweil's theories regarding AI, Transhumanism and the Singularity?
He paints quite a vivid picture of how things are going to come to pass and where we as humans
might fit into this scheme. Very interesting man and his ideas.

Again welcome!!

[Bragi]

Hi Tom,
17 he, nice age. Still plenty of time to get to learn everything you need!
Welcome to the forum!

[Freddy]

Hi and welcome Tom, good to see fresh ideas.

And yes plenty of time to learn things

[infurl]

17 he, nice age. Still plenty of time to get to learn everything you need!

You should try telling that to my 17 year old son. He's studying for his exams at the moment.

Luckily my office is some distance from the house, minimum safe distance that is.

[Bragi]

Ye, that's the downside of the whole age.   

[Data]

And a welcome to Ai Dreams Tom, from me 

[Kappers]

Have you read much into Ray Kurzweil's theories regarding AI, Transhumanism and the Singularity?
He paints quite a vivid picture of how things are going to come to pass and where we as humans
might fit into this scheme. Very interesting man and his ideas.

I have read about Ray Kurzweil's work as a leading futurist, and obviously his fleshing out of what we can expect from a technological singularity! Yet I haven't done great research beyond basic ideas and principles, I aim to do so. I have however read work by Eliezer Yudkowsky, he like Kurzweil has many theories regarding singularity and the future of AI as it challenges us humans. I'm guessing you have heard of his work? If not I would highly recommend it, on top of being a researcher at the Singularity Institute, he's quite reputable!

And yes plenty of time to learn things

I hope so! I want to be in a great position come university studies, and believe I'm on the right track currently. Although I'm struggling to decide where to start in terms of practical implementation of basic AI  . Because of this I've ordered "Artificial Intelligence: A Modern Approach" with high hopes.

Thanks for the warm welcome everyone, I really appreciate it (lots of forums are too daunting these days).
Tom