On 4/17/2015 6:58 AM, Thomas Johnston wrote:
> Is Watson rules-based AI or connectionist (neural network) AI? (01)
The IBM research project for the Jeopardy! challenge put together
a wide range of technologies for many different tasks that have
to work together. For any AI paradigm X, if you ask "Is Watson X?",
the answer is either yes, some versions are, or it could be. (02)
In any case, there are many issues involved: (03)
1. Rule-based systems and connectionist systems have very different
goals and solve very different problems. They are complementary
rather than competing. (04)
2. Among other things, IBM has also designed a chip with 5.4 billion
transistors that is specialized for implementing the so-called
"neural networks" -- which aren't really a realistic model of
actual neurons. For a brief summary of the technology, see
http://spectrum.ieee.org/computing/hardware/how-ibm-got-brainlike-efficiency-from-the-truenorth-chip (05)
3. You can't implement a complete system with today's version of
neural networks. Even the IBM chip can only serve as one
component of a much larger system. If it were plugged into
Watson, it could perform many functions quite well, but there
are many other kinds of AI tasks that would be better handled
by more conventional software. (06)
I discussed some related issues in the following slides: (07)
http://www.jfsowa.com/talks/micai.pdf
Why has AI failed? And how can it succeed? (08)
John (09)
_________________________________________________________________
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Config Subscr: http://ontolog.cim3.net/mailman/listinfo/ontolog-forum/
Unsubscribe: mailto:ontolog-forum-leave@xxxxxxxxxxxxxxxx
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To join: http://ontolog.cim3.net/cgi-bin/wiki.pl?WikiHomePage#nid1J (010)
|