Will be very interesting to see how this all turns out. It will surely depend on maintaining good relations between the libre and commercial partners. But can the greedies be kept at bay?
OpenAI, a leading machine-learning lab, has launched for-profit spin-off OpenAI LP – so it can put investors' cash toward the expensive task of building artificial general intelligence. The San-Francisco-headquartered organisation was founded in late 2015 as a nonprofit, with a mission to build, and encourage the development …
No, they cannot because they use the technology partially, only parse and not annotate.
Please be kind enough to open almost any word's definition in Merriam Webster? Do you see that the definition itself (body) is first given, and after it how the word is used?
But people from OpenAI try to replace the body with the second part only! (Read their blog (please)? They themselves said that they are trying to create a word's definition body through examples of how it is used.) Only when all a pattern's words are defined its definition body - only then the computer understands what is being said and is AI. In other words they should attach explanations-annotations, which is called in Math "tuples".
The present training of data is wrong, this way OpenAI compare contexts only and don't see-match subtexts - dictionary definitions are the subtexts and ignored!
Thus they should start to use the whole technology because otherwise it doesn't work. Read my patents.
OpenAI LP stated it clearly it wants to "raise investment capital and attract employees with startup-like equity."
Attractive employees with startup-like equity = Valuable Novel Proprietary Intellectual Property Holders.
Akin to OpenAI LP Tempting Black Box Brains ..... with SMARTR CyberIntelAIgent Operations/Virtually Real Programs for Presentation.
Money is well spent on those sorts of AIdVentures. The Returns on Investment are Stratospherically Epic. ...... and what else out there can guarantee you delivery of anything like that.
The greatest battles you will have are always with yourself as you fail to not simply believe virtually all shared above is so easily made when true.
It is not so.
As I said above, the guys work with contexts; that is they do parsing, structure texts and annotate the resulting patterns with other structured texts, similar to this. As a result, they annotate whole patterns, not the words that make them up. Plus they use only explicit information (training contexts), and completely ignore implicit (subtexts, the meanings of words from dictionary). This means that the computer can find other patterns similar to this pattern, not understanding what it is looking for and finding.
The way out is to annotate all the words according to their subtexts, that is definitions from the dictionary.
So the technology used by OpenAI is flawed and OpenAI in the present format has no future.
Yeah, sure. For the moment. Until the for-profit company's stock soars so high that eagles gasp for breath. Then there will be a decision that, regretfully, the duty of the board is to the well-being of its shareholders, so, off you go, little boffins, back to your science desks and let the big things be handled by the Men In Suits.
There is a reason Armani does not make lab coats.
Doesn't have to do everything a human can do to be useful. If it could do everything an 8 year old or a learning disabled adult could do it would still be a huge breakthrough (and cause a crisis for society/government) as it could do at least half the jobs that people do now.
Biting the hand that feeds IT © 1998–2019