Very cute, 9pla. I like him too

Jan Bogaerts - Mar 18, 2011:

Ye, AI has been over-hyped for so long, people stopped believing. In the mean time, when you look around, there’s AI (be it simple) all over the place: roomba’s, smart cars,....

A.I. computer programming, that works better than anything we see today, has been around years before the first personal computer.  Just look at industrial computers,  embedded computers, single board computers, micro-controllers, and computer museum machines that companies still have in service, because after all these years they still work ingeniously.

A.I. has been around before the first operating system.  for example, PL/M is a computer programming language by Intel from 1972, before the first personal computer.  It was used in office buildings, as elevator controllers, and to control oil burners, and is still in service today, running A.I. 

How do you think, when you press the elevator button, the doors open immediately, as if the elevator predicted you would be there.  It’s not just a convenience for traffic in and out of the building.  It saves wear and tear on the expensive elevator equipment which saves money. 

8PLA • NET - Mar 18, 2011:

C R,

What do you mean?

A.I. breakthroughs are happening all the time…  http://79b.org/A.I.breakthrough.php

- Just made this breakthrough today especially for you guys :- )

OMG, IT’S THE SINGULARITY!!!!!!!!!!!!!!!!!!!!!

Hans Peter Willems - Mar 18, 2011:

Victor Shulist - Mar 18, 2011:

My own 2 cents on the development of true AI is, it won’t be someone thinking up one single complex algorithm that will ‘do it all’...

Ahum, well, actually…...

Hans Peter feels real close….. ok… I could be wrong…. but for now, my money is on the step by step outlook.  It will be great if I’m proven wrong though !!

Hans Peter Willems - Mar 18, 2011:

Victor Shulist - Mar 18, 2011:

Narrow AI at least, has certainly been achieved.

Agreed, but it’s EXTREMELY narrow.

Agreed.

Victor Shulist - Mar 18, 2011:

Hans Peter feels real close…..

Well, there was a smiley for a reason… I was (somewhat) joking of course.

I do feel close, but not REAL close. I still have a boatload of work to do, I have a fairly complete base-model but now I have to start building on top of that. And besides that, the most important step is to validate my model against all the research information that is available in relation to my model.

Hans Peter Willems - Mar 18, 2011:

Victor Shulist - Mar 18, 2011:

Hans Peter feels real close…..

Well, there was a smiley for a reason… I was (somewhat) joking of course.

I do feel close, but not REAL close. I still have a boatload of work to do, I have a fairly complete base-model but now I have to start building on top of that. And besides that, the most important step is to validate my model against all the research information that is available in relation to my model.

Understood.

Then I have to admit I was also uninformed of the state of the art of robotics: I had not heard or seen anything about the humanoid robot Asimo, which somebody finally mentioned in this thread as an outstanding contradiction of the claims of that book author (Stephen Gislason). I have to agree about Asimo: I looked up Asimo on YouTube and found some very impressive demos, especially of two-footed running, which impressed me the most—I didn’t know that was possible with the state of the art…

new version amazing robot asimo
http://www.youtube.com/watch?v=Q3C5sc8b3xM

ASIMO - All New Features 2011
http://www.youtube.com/watch?v=zul8ACjZI18

The grasping of objects such as trays and human hands also impressed me. Then I got to the videos that weren’t edited promotional materials and found the embarrassing failures of Asimo while in front of an audience…

Poor Asimo Falls
http://www.youtube.com/watch?v=cZeseJBzuF4

oops Asimo screws up
http://www.youtube.com/watch?v=2FDeEuH6mCQ

I had to laugh at both of those disgraceful failures. The promoters sure didn’t know how to handle a failed demo, either: rather than dropping a curtain or putting up a privacy screen to hide their failure, they should have had a backup plan to provide humor, like to send in a couple guys dressed in white with a stretcher to cart the thing off, then maybe replace it with a miniature humanoid robot like Qbo or Nao…

QBO Robot in front of a mirror ( UPDATED )
http://www.youtube.com/watch?v=TphFUYRAx_c

NAO demonstration 1
http://www.youtube.com/watch?v=qkWoRtD5R14

I also noticed that the kitchen tray demonstration of Asimo was basically a laboratory set-up, with a pure white tray, pure white countertop, and clearly marked diagram on the counter where the tray was supposed to go. That means Gislason was largely correct…

“For now, no robot is adaptable enough to work in home kitchens that humans find easy to use. To use a robot you have to build a robot kitchen.”
http://www.nutramed.com/Philosophy/digital_limits.htm

So basically we’re still dealing with the laboratory microworlds / toy worlds of the early 1970s that David Marr described as having been a dead end per se…

The second approach was to try for depth of analysis by restricting the scope to a world of single, illuminated, matte white toy blocks set against a black background. The blocks could occur in any shapes provided only that all faces were planar and all edges were straight. This restriction allowed more specialized techniques to be used, but it still did not make the problem easy. The Binford-Horn line finder (Horn, 1973) was used to find edges, and both it and its sequel (described in Shirai, 1973) made use of the special circumstances of the environment, such as the fact that all edges there were straight.
These techniques did work reasonably well, however, and they allowed a preliminary analysis of later problems to emerge—roughly, what does one do once a complete line drawing has been extracted from a scene? Studies of this had begun sometime before with Roberts (1965) and Guzman (1968), and they culiminated in the works of Waltz (1975) and Mackworth (1973), which essentially solved the interpretation problem for line drawings derived from images of prismatic solids. Waltz’s work had a particularly dramatic impact, because it was the first to show explicitly that an exhaustive analysis of all possible physical arrangements of surfaces, edges, and shadows could lead to an effective and efficient algorithm for interpreting an actual image. Figure 1-3 and its legend convey the main
ideas behind Waltz’s theory.
The hope that lay behind this work was, of course, that once the toy world of white blocks had been understood, the solutions found there could be generalized, providing the basis for attacking the more complex problems posted by a richer visual environment. Unfortunately, this turned out not to be so. For the roots of the approach that was eventually successful, we have to look at the third kind of development that was taking place then.
(“Vision: A Computational Investigation into the Human Representation and Processing of Visual Information”, David Marr, 1982)

So I believe Gislason wasn’t that far off.

Maybe somebody should post in this forum a summary of the state of the art in humanoid robots, something like the following?

World’s Top3 Humanoid Robots - Asimo vs HPR-4 vs NAO!
http://www.youtube.com/watch?v=N_m56irWKeI