Slashdot Mirror

This sounds very much like the guessing game done as part of the talking heads experiments in 1999 at the VUB Artificial Intelligence Lab and Sony Computer Science Lab Paris by professor Luc Steels. These experiments already date from 1999.

This sounds very much like the guessing game done as part of the talking heads experiments in 1999 at the VUB Artificial Intelligence Lab and Sony Computer Science Lab Paris by professor Luc Steels. These experiments already date from 1999.

See also here for previous work on this idea: http://www.csl.sony.fr/items/2001/musaicing/

For some prior art, see also: http://www.csl.sony.fr/~pachet/Continuator/index.html

So robotics as part of AI is indeed a very active area for research.

The enhanced AIBOs used in the experiments have defined inputs (perceptions) and defined actions (outputs). The best way to think of this is to pretend there's a "little man" (homunculus) inside your brain: he only see the inputs from your eyes and controls the output signals to your limbs. Then replace the little man with a computer program. Yes, the computer program can only do what it is programmed to do, but as "Good Old Fashioned AI" has shown, the program can also can alter itself, so there is no lack of variety in resultant behaviors. The key is to choose an appropriate set of initial behaviors for the particular environment.

You should skim some of the papers of Luc Steels, who is the primary researcher behind this work. Don't dig into the details; look for the summaries and the insights. Steels is a good writer and has gone out of his way to encourage research. Here are two sources:

• Sony's CSL Labs publications - go down to Steels' papers
• Luc Steels publications.

A favorite, How to do Experiments in Artificial Language Evolution and Why. For an example that addresses your question, scroll down to Figure 3 which shows the AIBO and a plot of what it "sees" with it's rather primitive visual system.

The enhanced AIBOs used in the experiments have defined inputs (perceptions) and defined actions (outputs). The best way to think of this is to pretend there's a "little man" (homunculus) inside your brain: he only see the inputs from your eyes and controls the output signals to your limbs. Then replace the little man with a computer program. Yes, the computer program can only do what it is programmed to do, but as "Good Old Fashioned AI" has shown, the program can also can alter itself, so there is no lack of variety in resultant behaviors. The key is to choose an appropriate set of initial behaviors for the particular environment.

You should skim some of the papers of Luc Steels, who is the primary researcher behind this work. Don't dig into the details; look for the summaries and the insights. Steels is a good writer and has gone out of his way to encourage research. Here are two sources:

• Sony's CSL Labs publications - go down to Steels' papers
• Luc Steels publications.

A favorite, How to do Experiments in Artificial Language Evolution and Why. For an example that addresses your question, scroll down to Figure 3 which shows the AIBO and a plot of what it "sees" with it's rather primitive visual system.

Despite the generated jokes about dogs and the French, and the "oohing and aahing of the crowd at the AIBO robotics soccer games broadcast on U.S. national television, this is not merely "cute". This may be the most important research that you have ever read about.

Researchers Luc Steels and colleagues at Sony's Paris Computer Science Laboratory in France have performed a series of remarkable experiments demonstrating the development, from naught, of spoken language among robots. Words, grammar and semantics evolve spontaneously among cooperating robotic agents initially programmed with a small base set of ground perceptions and behaviors. And from the development of language arises cooperative group (intelligent) behavior.

Enhanced AIBOs are initially programmed to recognise simple stimuli from their surprisingly limited hardware sensors. Over the course of several hours or days, the AIBOs learn to distinguish objects and how to interact with them. A built-in curiosity system ('metabrain') continually directs the AIBOs to look for new and more challenging tasks and to cease activities that are not fruitful. In time they develop more complex tasks, just as do human children.

Like children, the enhanced Sony AIBOs initially babble ("argue?") until two or more settle on a sound to describe an object or aspect of their environment. Over time the group gradually builds a lexicon and grammatical rules through which to communicate. Agreement on word usage spreads through the population as terms for similar meanings compete for acceptance. For example, the robots develop the language structures to express that a red ball is rolling to the left. Just as human twins sometimes develop a unique language in which only they can communicate, the enhanced AIBOs (which are clone-like and similar to twins) develop their own language.

Language analysis and generation are part of Good Old Fashioned AI (GOFAI) and have been studied extensively for decades by AI researchers. In the past several decades GOFAI was challenged by Nouvelle AI (Situated AI) championed by Hans Moravec and Rodney Brooks. This alternative approach holds that true AI will not arise from formal mathematical systems but instead from robotic behaviors which have a subsumption architecture as an overall organising principle for the individual robot. This architecture consists of layers of behavioural modules, each capable of carrying out a complete but simple task. Steels' enhanced AIBOs are embodiments of just such a subsumption architecture and provide strong support for Moravec's and Brooks' hypotheses

Prior to Luc Steels' experiments, no one had experimentally demonstrated how language develops among intelligent agents. Steels' experiments are no less than stunning: in a controlled environment AIBO robots develop their own words and grammars for objects in their environment. All aspects of human language development are mirrored in these experiments: words compete for acceptance in the population, new words are created, and grammatical structures arise spontaneously. Steels' work also addresses the idea of a "robot culture", since it is in the context of a population of cooperating agents that language becomes most useful.

Contrast this with the W3C's Semantic Web effort, which has received much more interest and money in recent years due to the growth of the Internet yet has proven far less fertile. In the Semantic Web there are multiple competing "ontologies" (roughly, data dictionaries wherein all terms are strictly defined by specialists from their

Information about the actual AIBO setup:
http://www.csl.sony.fr/perspective/
Videos here: http://www.csl.sony.fr/perspective/node6.html

Information about the actual AIBO setup:
http://www.csl.sony.fr/perspective/
Videos here: http://www.csl.sony.fr/perspective/node6.html

This type of language/vocabulary development experiments has been done before.
You should take a look at the talking heads experiment.
This page has some related publications.

You didn't read the complete article ? Sony went after the hobbyist who made the Aibo dance...

Well he is back up and running now and he is Legit-i-Mutt

I think that Sony didn't like the fact that the whole aibo code was online, competitors could then steal it ect. The lawyers have to justify their jobs as well.

The fact that he is back online now does show that Sony are not entirely unreasnable. They actualy use modded Aibos themselves. They have one which can be trained with a clicker

I think some companies are starting to wake up to the fact that modding can be a good idea. It can prolong the life of a product. It can also come up with novel uses, which can then be re enjeneared and put into the next version, for profit

There is a flip side. Copywrite issues do piss companys off but more importantly if Mr Smith buys a product and it gets modded with some buggy software. Mr Smith may think that it is the product that is buggy. How may PSXs' died cause the mod chip burned out? This was not Sonys fault but if people didn't realise it was the chip rather than the legit parts it may make them more reluctant to buy another sony product

BBC News is now reporting that Sony researchers are experimenting with increased speech capacity. Here's one of the first papers about the increased-capacity talking Aibo project. This is the English/HTML translation of the French/PDF version (which seems to be unavailable for download) and so is a little messy. Unfortunately, Sony's Computer Research Laboratory seems to be down at the moment. As an anthropologist interested in the evolution of speech, I'm absolutely fascinated by this, and whish I knew where to find more of the speech recognition software specs.

Do androids dream of electric sheep yet?