Image courtesy MIT-IBM AI Lab |
MIT and IBM founded a joint research lab for AI[1] research
one year ago. Normally, we tend to ignore such announcements as past such
corporate-academic deals often disappear without a trace.
Therefore, generally
it’s safest to ignore the initial announcement and wait for actual results.
However, IBM offered a briefing on the lab, and the company has other
interesting research activities underway. Therefore, we accepted the
invitation; taking the chance that there might be real substance behind it.
|
If
you are interested, Oct 1 - 5 is Inaugural AI Research Week at the MIT-IBM
Lab in Cambridge, MA. Many events, but not all, are private and by
invitation only. Find out more here: https://tinyurl.com/ya7zahev.
|
Director Cox provided additional details on how the lab will
function. IBM has committed $240 million to the project. The project has a
timeline of 10 years. A total equivalent of 100 full-time researchers with 60
from MIT and 40 from IBM will work on the projects. In our opinion, there exist few companies
able to stack their research departments against MIT, one of the leading
research institutions in the world. Kudos to IBM.
From what we heard, great benefits ought to accrue to both
MIT and IBM in this partnership. Any agreed upon project will have a multiplier
effect as top IBM researchers and MIT people will be working on it. IBM brings many strengths to these efforts,
e.g. contributing the latest working technology and equipment. From MIT, the intellectual resources that it
has available would be hard to match.
Of course, none of the above will guarantee success. Many
research projects fail. It would be a serious mistake to demand quick results,
or even results that can be incorporated into product development. Also, as we
understand it, any results will be published and tools etc. will be available
in open source. There appears to be no effort to create IBM proprietary results.
Instead, results will focus on advancing the overall AI community.
Director Cox’s briefing contained other interesting points
including details on several of the 48 projects. In each case, he identified
specific individuals from both IBM and MIT who would be working on the project.
When questioned for more details on several projects, he responded it was too
early for details. He made no attempt to invent anything. As noted above, he is
a very serious person. We think that IBM made an excellent choice in recruiting
him.
One research area that did stand out was the idea that
ethics should play a significant part in AI. In today’s world where we see a
major effort by the Chinese government to use AI to control its people we can
only applaud the effort to look for practical examples for the use of ethics in
the design of AI systems.
Despite our initial prejudice, we’re convinced this is a
serious effort by IBM and MIT to benefit the broader AI industry. It provides a
very nice model for cooperative partnerships between business and academia. We
hope it produces the type of results planned, and patiently look forward to
them. A side note, we hope we will be able to understand them! Congratulations
and good luck to IBM and MIT.
No comments:
Post a Comment