I’ve been way to busy helping different groups wring insights from GBD estimates with microsimulation. But in my copious spare time, I’ve been enjoying a paper that studies what factors allow different groups to use models.
The authors focus on a particular Systems Dynamics model of the health system, but I think their finding can be generalized to other models and other systems. For example, “the actual utility of the model is dependent upon a match between the scope of the community’s defined problem and the capabilities of the model.” That makes sense to me, but I’m not sure I thought of it explicitly before I saw it written out.
I sat-in on a CSE seminar recently, where a big crowd is exploring the state of the art in human-and-computer-together intelligence. It was really fun. The topic was a discussion of a paper on human/computer collaboration from the 1990s:
But just as fun as the classic article and discussion it inspired, was an even older vision of what digital assistants might be, from Apple in the 1980s:
I left thinking that a knowledge navigator like the one Apple envisioned is not really collaboration, but when it makes the Brazil and Sahara simulations work together, that might be collaboration. But to be a true collaborator, both agents need to want something (or “desire” something?) for themselves.
I hope I have time to attend again soon.
Comments Off on Augmented Intelligence: what is collaboration?
I read random papers once in a while from the AMS Math Reviews program, and I read one recently about an MCMC approach to X-ray imaging. It was a fun, detailed look at a few different ways to do sampling, and use effective sample size to figure out which worked better when.
It did also leave me wondering what the giant X-ray machines buried 1,000 feet underground are for, though.