City of Amnesia

Ok, so a somewhat complicated robotic idea.

What if I make a chat bot, save the template and upload this into a robotic body? Give this robot an expandable dictionary in it's mainframe so that it can learn new words and teach it speech patterns.

Next give it the ability to walk, and pick things up/ move them. Give it a personality and see what happens from there.

except the walk/ move things everything else is not too complicated, and I'm hoping to have this done in 5-6 years (have to be realistic here).

My thinking is to make an actual R. Dorothy 2, seeing as she is only 4'11" and we have humanoid robots in the 5' range building the body might not be too hard, but programming the personality and getting it to do what I want may be the issue. I wouldn't be working alone, but the thought of doing it seems like a good one.

So what is the probability of doing this in 5-6 years? The chat bot and ever expanding speech and dictionary might not be hard, but incorporating it might be.

Welp

(http://www.starstore.com/acatalog/terminator-l-sideshow.jpg)

(Oh and I friended you)

Erm what? You friended me? You mean on FB?

And why does everyone worry about true AI? it won't happen for at least 100 more years.

...

Good luck.

Any updates? I'd love to hear how it's coming!

don't have any updates yet. I'm still trying to find a good chat bot that I can program to do what I want. It would also have to allow for visual effects.

I did figure out that making a small body wouldn't be too difficult, but it would only be 3" tall then. Pocket Dorothy might be fun, but would scare the living shit outta most people I know. And it would need to be hooked up to a computer most of the time. So not a very good option as of right now.

I've not been able to give much thought to the project because of school, but I'll have a break soon so I an look into it then.

It's possible. Youtube had an AI robotics project from Hanson Robotics called "Jules":

• http://www.youtube.com/watch?v=elVdiboCa80
• http://www.youtube.com/watch?v=ysU56JzBjTY
• http://www.youtube.com/watch?v=xRR33WDFi_k

He is now a guest of Bristol Robotics Laboratory: www.brl.ac.uk/projects/empathy/heads.html

He kind of vaguely reminds me of Ewan McGregor, for reasons I can't describe...

Here also is Hanson Robotic's Wikipedia page:  http://en.wikipedia.org/wiki/Hanson_Robotics

I doubt you're going to start with this level of complexity, but just showing you that it's possible. No worry about a Terminator, at least without an ANN (http://en.wikipedia.org/wiki/Artificial_neural_network). Maybe, instill Asimov's 3 laws? Also, a pocket Dorothy that would walk to the edge of a table and lean precariously over the edge would be awesome!

P.S. I didn't friend you, but I ran across "Yuko" references on the "The web". Interesting...

Hm? what do you mean cross reference the web? If you where looking for a fb page you won't find it that way.
And there is no way I'd be starting at that level. Just building a Pocket Dorothy would take six months. 3 1/2 if I had help from a pro.

WHY ARE SO MANY PEOPLE TRYING SO HARD TO MAKE TERMINATOR HAPPEN?

Well, technically, one of the search engines does show fb pictures, but it probably only shows unlocked ones. However, no, I meant references to your avatar "Yuko-san" and her stories. Interesting choice of avatar.



Sorry, I didn't mean to overwhelm you with advanced projects like that. Wow, 3 1/2 to 6 months?!? That quick?

 I guess you could isolate the motion control software/hardware from the chat bot software/hardware in the beginning. So, it would walk and talk, but not react (ex. you couldn't ask it to walk over to you). However, I guess you could have some sort of command software on the motion control and two microphones (one to the chat bot, one to the motion control) in the design. That way, if you say ("Hey, over here!"), the motion control software "hears" and sets R. Dorothy walking over to you, while the chat bot software hears the same comment (via its microphone), and has R. Dorothy say something appropriate (ex. "Hello... Have you seen Norman?").

^that is not a bad idea at all. I'll go download the chatbot I was thinking of using.

The reason the body wouldn't take that long is I have scrap metal and other various parts that can be put into a small automaton lying around. Plus I have an old dinosaur robot toy that I never built, who's small engine could be used in a 3" Pocket Dorothy. I would have to mod the engine a little, but it shouldn't be hard.

BECAUSE TERMINATOR IS FROCKING AWESOME.

Dude.

Also, as unusual and undoubtedly difficult this project is and will be, Yuko, good luck.

^ I second that.

However, if it's female and made of scrap metal and spare parts, I would say "she" will be more like:
(http://upload.wikimedia.org/wikipedia/en/thumb/d/d7/BAA_1_cover.jpg/230px-BAA_1_cover.jpg)
and Alita is always good as gold...

No matter what, I'm sure "she'll" be awesome. Good luck on your project, Yuko!

Waiting for an update.

I'm waiting till this fall to actually start programing. That programing course I took in college sucked cause we used Alice 2.2, and if you ever try to use it you'll be as confused as I was. I went into that class knowing how to program in DoS box and using the command line. I came out of it knowing nothing more then I hated object oriented programming. Theres a teacher who teaches python and C++, so I'm going to take him and then program the chat-bot.

As to building tiny robots, my favorite site to find robot parts went down a few weeks ago and isn't back up yet.

Moving from purely procedural languages to object-oriented programming can be a bit jarring (I hope you're not used to goto), but often much more useful. That is, unless you're planning to learn COBOL and go into legacy software support, but I've heard there's actually some demand for programmers with those skills.

I've been working with mostly dynamic scripting languages lately, though not Python. Never used Alice, but just reading the blurb on the home page makes it sound terrible if you've had any previous experience programming. Good luck with Python/C++.

Best. Thread. Ever.

Some time earlier this year, I was struck by a most interesting realization. I have scripted and coded IRC AI bots for some years. When I say AI, I obviously mean something that imitates humans rather than anything sentient. Additionally, the vast majority of things I have put together are nothing more than intricate trolling tools (such as bots designed to imitate the behaviors of schizophrenics and rant 24/7 for days on end). My realization was that, however basic these things may be, that if I could connect them to hardware I would have real robots. My problems are as follows: I don't have the money for motors or other parts, nor do I have the level of programming knowledge to connect a program to hardware, nor do I have the motivation to sit down for a few hours and learn (I'm a lazy bum).

Two friends of mine (Zola from PCF and her roommate) have both the technical know how and the dedication to produce at least prototype androids, but they do not presently hold the resources for any major projects. I followed their work obsessively as a child and as a teenager, and as an adult ultimately drew the conclusion that the world has made a terrible mistake in neglecting such geniuses and that it's a lack of investment, not technical problems, which has prevented the emergence of androids.

As for you, Yuko-san, I wish you well with this project. I would like to make a recommendation: have the robot record everything it sees, everything it hears, everything it does. I do not believe it will take a century for us to learn how the human brain works and recreate it. Your creation, if it were to be upgraded to sentience in the future, would thus have an extensive memory on which to base its identity.

Maybe when I'm done with trig and programming this semester I will have a small chance to do this. My plan has always been to build a small childsized robot if I got the chance. I have some parts that could be used (scrap metal and old comp parts are a poor robotists best friends), it is merely time and programming knowledge I don't have as of yet.

Also ebay is a good place to get some parts, as well as my friend that has a small computer fixing store. He has tons of old parts I could get and use.

Yuko, I demand you start this project! It must be done.

I need to, but I changed from robotics to programming as my major. Which will make it easier to program it I guess. I can build small robots but I'll definitely need help. Like making a body chase for the animatron to fit in. I'll get started on the programming when my school work load lightens up.

Still interested in hearing about this.