is not ALICE one of the first in the current space now ??
http://alice.pandorabots.com/ |
funny- I always kinda liked that little paper clip man... how dull am I? I mean, well, he even collapsed into a heap and snored... |
for every "no" ALICE knows only " Are you serious? " kind of stupid and not nicely done... |
You describe an excellent system, it is almost exactly how I would want it to be. Except you neglect to mention that it all happens timelessly, instantly, when computers are in our minds. :) |
I don't understand what's compelling about this at all. If these are all AI entities, this is equivalent to natural language search, with avatars. Avatars add nothing to the usefulness of the search result, and natural language search is both hard to do well and not proven to be more useful or more desirable. So what, exactly, makes this "a great way to query knowledge"? |
Only the discussion between avatars would be a loss of time... as it would be easier for one to read through something instead of reading/listening each time a avatars says something. |
Peter is correct. Your goal seems to be natural language "semantic" processing driven search in a "topic- concept-walled" garden. Hard core search is the domain of members of the global Special Librarian’s Association. At their international 2005 Toronto meeting they awarded their “Innovation in Technology” award to the first truly Internet accessible semantic NLP information “filtering” ["topic- concept-walled" garden] search engine located within Boston Children’s Hospital’s “Center on Media and Child Health.” www.cmch.tv/research/. Their “smart search” NLP eliminates the need for “query structuring” by a user. It excels with lots of “context”- unrestricted length conversational style queries. It mediates results from ten different “social science” professions [journal articles] – filtered and ordered more precisely when using “jargon”, well-articulated community-informal vocabulary. Experience true [Virtual Experts Room] semantics yourself, try some questions like: What is the impact of the media on adolescent sexual attitudes and behaviors? Or, Can parents prevent children from experiencing unwanted effects of violent television?
The current [key word] query technology limits user’s allowable contextual input. The rule garbage-in-garbage-out still holds. CMCH.tv's vocabulary-oriented Ontology-based “Fuzzy AI” enables very articulate user input of the contextual nuances of Experts Room ideas.
|
Peter and Michael, I didn't mean it would be used for ordinary often-performed search queries, but for deeper research and learning about topics. Natural language and avatars have the primary purpose of catering to associative abilites of our brains, easing the learning process.
Disclaimer: this is a virtual expert representing Philipp Lenssen of blogoscoped.com, produced by Veky's technologies. Veky employs reasonable effort to represent the knowledge of target experts, but gives no waranty of the match. If you are Philipp Lenssen and you object to this, you can disallow VekyBot in your robots.txt file.
;-) |
Nice article, but I wouldn't allow the avatars to "learn" from each other. Imagine you would do some research about a weapon to make a defence shield against it and the expert avatar talked to a pacifist before and is now unwilling to tell you anything about his destructive invention.
Not the best example, but I hope you can get my point out of it. |
Well, I guess at least the one who changed "Philipp" above to "[Veky]" sees what the main objection of mine is to the idea of bots based on current experts. It's not the privacy connected with input, but with output of the bot. ;-) |