Slashdot Mirror
IBM Opens Up Its Watson Supercomputer To Researchers
An anonymous reader writes IBM has announced the "Watson Discovery Advisor" a cloud-based tool that will let researchers comb through massive troves of data, looking for insights and connections. The company says it's a major expansion in capabilities for the Watson Group, which IBM seeded with a $1 billion investment. "Scientific discovery takes us to a different level as a learning system," said Steve Gold, vice president of the Watson Group. "Watson can provide insights into the information independent of the question. The ability to connect the dots opens up a new world of possibilities."
My homemade chatbot has the same problem.
Fifty years of Yippie! 1968-2018
To use it, contact IBM, they in turn will send 'engineers' (really business sales men) to discuss it with your boss (not with you, you are too technical and can see through their poorly written language parser).
Those 'engineers' will try to put in lots of 'consultants' from IBM to interface their revolutionary new parse onto your data at great expense. Those will demoralize and undermine your programmers to try to take over the role in the company.
When they deliver something... eventually..., they'll then market it as a huge success and your boss will pretend it was, because he doesn't want to look like an idiot. IBM will continue to milk maintenance money from the company bleeding it dry with comically incompetent support staff.
Boss will leave to join IBM's team of 'engineers' perhaps.
Excuse my negativity, but IBM does not permit public comparison of its crap technology, and anyone who has benchmarked an IBM mainframe knows how big the gap is between their claims and the reality. The product here isn't 'Watson', it's IBM consultancy, which in my book has a negative value associated with it (based on a previous experience of IBM infesting a corp). Watson is just a marketing exercise used for novelty value.
Why is the term "Supercomputer" being used to describe Watson? No demonstrated systems have shown anywhere near the processor or node count that actual supercomputers have (the Watson machine on Jeopardy for instance was only 90 nodes with around 2K cores). Also it uses an off the shelf interconnect (10gbit fiber) with a simple hierarchical network fabric which doesn't even approach even small supercomputers in terms of performance (which use something like Infinband or Seastar in a N-Dimensional torus interconnect topology).
While I have nothing against the technology being used for Watson. The fact is that it is not a supercomputer and the division of IBM that did make supercomputers (BlueGene) has been disbanded (with most of the key individuals leaving for other places).
In related news:
“To put this in perspective with p53, there are over 70,000 papers published on this protein. Even if I’m reading five papers a day, it could take me nearly 38 years to completely understand all of the research already available today on this protein. Watson has demonstrated the potential to accelerate the rate and the quality of breakthrough discoveries."
Using [Watson], Lichtarge’s team identified proteins that modify p53, which is a key protein related to many cancers. Cancer researchers usually only find around one new protein to work on a year, but the Watson collaboration discovered six potential proteins to target for new research, according to IBM.
a cloud-based tool that will let researchers comb through massive troves of data, looking for insights and connections
Sort of reminiscent of BACON, isn't it? (With more intelligence, I presume.)
Ezekiel 23:20
For any R&D company that has a lot of in-house raw data, the Watson Discovery Advisor is likley to generate a lot of interest.
Imagine you're an executive VP in R&D in a board meeting. You receive this challenge from the CEO who hates your guts: "Our R&D productivity continues to decline. What're you doing about this? How are you extracting every last bit of value from our data? Our major competitors are using tools like Watson. Why aren't we?" You damned well better have an answer.
I work for a Fortune 100 R&D company that is *very* interested in improving its R&D ROI. I know for a fact that any opportunity to reevaluate our data to derive additional value (e.g. new prospects) will set off bells among the C suiters. IMHO Watson, and especially Discovery Advisor, is the first system I've seen with that potential.
Of course, IBM is going to have to step up its game in loading and tagging all that data. I suspect that's where most of its new Watson staff will work. I suspect the most fruitful features in data are not readable in natural language (English). Much has been summarized in graphs, or lies in tables, or in addenda. Or it's buried deep in old screening results stored in flat files that were long ago archived to tape. And it's certainly not present in easy-to-access content like online research paper abstracts.
But all it takes is one or two significant new leads to make the millions you spent hiring Watson look like money *very* well spent. And personally, I think that scenario is entirely plausible.