[doanviettrung]

Please read on if you share my beliefs that not only humans but some animals are conscious too, and that we can make some machines conscious.

WHAT WE WILL DO: TOGETHER WE'LL LEARN THEN BUILD A CREATURE

I am not a professional consciousness researcher, but I have spent a few hundred hours reading research papers and books. This learning convinces me that while we can't make a machine whose consciousness is as complex as humans (next century, perhaps), we can make a simple conscious machine now.

If you are interested, then I will share my ideas with you in detail about my proposed architecture, and perhaps you share yours. Then, each of us continues to read books and research papers to keep on learning and developing the architecture further. When we are ready, we write the specs for the software that implements the creature and its environment. Then, either you write all of that software, or we get some help. If we have hardware skills, we can try making the creature in hardware too.

BUDGET: YOU GET BETWEEN $0 AND $10,000 USD

I prefer that you don't do it for the money, rather we are partners attempting to tackle the machine-consciousness challenge. But because I want to be your fair partner, therefore if you do more work than I do (because I don't program), then I can use my own money to pay you up to 10 thousand dollars.

Remember, even $10,000 is not much compared with the hours we'll put in, and nothing compared with our satisfaction of scaling Everest, or the potential money we could earn if we find a way to monetise our results.

OUTLINE: A WEBSITE WHICH SHOWS OUR (VERY SIMPLE) CREATURE AND ITS THOUGHTS

So that you have a more concrete idea of what's involved, let me briefly outline:

- We will, via software, create a virtual environment which is very simple. Perhaps it's just a flat landscape, with just a "tree" which emits sounds for our creature to listen to and interact with.

- Our creature will also be extremely simple, it certainly won't be fury or talk or jump around. The simpler we make it, the less programming we have to do, because a complex body feeds a lot of information to its brain.

- If you agree, one of the ideas that we will use to architect our creature's brain will be Antonio Damasio's. Simply explained, Damasio theorised that the brain creates a first-order model which comprises all the body's signals (the organs, skin, ears, etc.). The brain also creates a second-order model, which encompasses the first one. Here, the brain sees the body interacting with the environment, and the sense of "self" arises from that.

Damasio may or may not be right about how the human brain develops the sense of self, but regardless, we can use this idea to help build our creature's brain.

- We will make a website. There, visitors can see our graphic representation of the creature, its environment, and also the stream of (very simple, very primitive) thoughts going on in our creature's brain. Hopefully, some of them will collaborate with us.

[modi123_1]

Wow.. sounds like grad school all over again...

[RudiVisser]

Even reproducing a SCO would take a year to do :|

[sparkart]

Sounds interesting and something worth the effort.

[virgul]

Although im not sure of how realistic of a project this is im actually quite interested in it.

[nooblet]

Google and their legion of 20,000+ engineers worldwide have been tackling the self aware and conscious artificial intelligent computer conundrum for years now. A meager 10k budget plus countless of hours of work from a few programmers on this forum won't really put a dent into what you're trying to achieve. There are already thousands if not hundreds of thousands of small machine learning programs people have built and entire courses on the subject.

While it is interesting and to some extent can be accomplished on a small scale, there are already a ton of people working on this. I don't see you proposing anything new.

[doanviettrung]

virgul, on 07 May 2010 - 05:09 PM, said:

Although im not sure of how realistic of a project this is im actually quite interested in it.

Hi Virgul

For a bit more info, please check out these articles of mine on conscious-robots.com:

- a 500-word article on Damasio's theory on how the human brain creates the mind
http://www.conscious...sciousness.html

- a 3500-word article on Haikonen's architecture for a conscious machine
http://www.conscious...scious-mac.html

- prof Haikonen's article commenting on the above
http://www.conscious...article-by.html

This post has been edited by doanviettrung: 09 May 2010 - 05:32 AM

[doanviettrung]

nooblet, on 08 May 2010 - 11:52 PM, said:

Google and their legion of 20,000+ engineers worldwide have been tackling the self aware and conscious artificial intelligent computer conundrum for years now. A meager 10k budget plus countless of hours of work from a few programmers on this forum won't really put a dent into what you're trying to achieve. There are already thousands if not hundreds of thousands of small machine learning programs people have built and entire courses on the subject.

While it is interesting and to some extent can be accomplished on a small scale, there are already a ton of people working on this. I don't see you proposing anything new.

Around the world only a few groups are trying to create a conscious humanoid robot. None has succeeded. There's no guarantee that we will, but our approach is different, and more do-able, because we try to make much simpler creatures. For example, they have no eyes, therefore we don't need to tackle the vision thing. As far as I know, no-one else is using this approach (ie. simple creatures + collaborative work + online access).

Even if we don't succeed, our learning along the way might create benefits, such as games whose characters are actually motivated to stay alive.

This post has been edited by doanviettrung: 09 May 2010 - 05:38 AM

[calvinthedestroyer]

I'd do it for 10k a month. I wouldn't bag my head against the wall for anything less.

[Motoma]

I'm interested. My screenname at gmail.

[SymphonyX]

Hey, I am very interested and sound something worth the effort.
Do you have an email address I can get in touch with you?

SymphonyX

[doanviettrung]

SymphonyX, on 09 June 2010 - 09:15 AM, said:

Hey, I am very interested and sound something worth the effort.
Do you have an email address I can get in touch with you?

SymphonyX

Hi SymphonyX, sorry I am slow in relying. My contact details are below. I look forward to hearing from you.

Trung Doan
Mobile +61 400 466 848 (I am in Melbourne, Australia)
Yahoo trocdaumel
Skype trocdaumel

[jcremeans11]

I'd do it, but I don't want to end up succeeding and the world turning into iRobot

This post has been edited by jcremeans11: 15 June 2010 - 12:11 PM

[modi123_1]

jcremeans11, on 15 June 2010 - 01:09 PM, said:

I'd do it, but I don't want to end up succeeding and the world turning into iRobot

Then fail good sir... fail as hard as you can.

[bcranger]

How simple is the creature to be?

I mean programming something to respond to various sounds that are familiar would be doable in an hour for just the software.