Back in September, Matthew Barakat of the Associated Press interviewed us about DAGGRE. His article ran today. Here’s an instance in the Boston Globe. CBS News picked it up for a radio interview, which presumably will run soon. (Hope I did OK.)
We’ve had several emails from interested participants: please sign up at http://daggre.org. We’re especially interested in people who really like to follow current events, and want to get better at forecasting. The better you do, the better we do. Ideally, participants would revisit their estimates weekly, and continue participating for at least a year.
Note: If you are in another one of the IARPA ACE projects, or plan to be, you cannot join DAGGRE. That’s IARPA’s rule, not ours.
In general I liked Barakat’s AP article, but I should make some specific replies.
“Your educated guess may be just as good as an expert’s opinion.”
–> That’s not quite right. My educated guess is not as good as, say, Henry Kissinger’s. But my guess plus hundreds of others from a diverse and uncorrelated crowd may well be better. I’d be happy to test, if Kissinger is willing.
“They are competing with four other teams led by professors at several universities. … At stake is grant money provided by the Intelligence Advanced Research Projects Activity”.
–> Again, not quite right. To be sure, there is healthy rivalry, but the IARPA Program Manager (PM) has gone out of his way to assure us that there is no mandatory downselect. I can’t speak for IARPA, but if everyone is producing good results, in theory all teams can continue past year 1. Of course, if you think that IARPA’s budget will get slashed, that’s different, but the assurance from the PM means there is much more potential for collaboration and doing better research. One of the exciting things about this program is that the PM has strongly supported the research aspect — we all want to find out what works and what doesn’t, and why. Ideally you want just enough rivalry to make us competitive, without the pressure to keep everything proprietary and thus stifle progress.
I liked this quote from IARPA’s spokeswoman: ‘”It’s all about strengthening the capabilities of our intelligence analysts,” Montgomery said.’
But to clarify: right now we’re doing research, not operations. We hope to demonstrate techniques that will later be applied to improve actual analysis. The goal for now is to find what works, provably. By assuming the role of neutral arbiter for a large pool of shared questions, IARPA is providing a rare opportunity to do just that.