5 ms·
MIT researchers find memory capacity much bigger than previously thought
- ilamont 18y agoInteresting article, but there is no actual size estimate. I wonder how this impacts Kurzweil's "Singularity" timeline?
- dangoldin 18y agoTo play devil's advocate here - maybe it's because many of their subjects were MIT students.
- Zev 18y agoIt says the research was done by MIT students, not the test subjects were MIT students. Perhaps they found students from other universities in the area or the general public?
- ilamont 18y agoEven among the general public in that area, the concentration of college students and grads is high relative to the rest of the country.
- akd 18y agoI've run psych studies at MIT. 90% of subjects are MIT students, unless the prof specifically has to get a broader cross-section in which case the study becomes annoying, expensive, and delayed. All general social science is implicitly the study of college students.
- Tichy 18y agoEven assuming that MIT students are way smarter than the average population, what factor would you consider? Do you think MIT students on average are a thousand times smarter than an average person? Or maybe just 3 to 4 times? Even with a 1000 times, the study would probably still show amazing capabilities for an average memory?
- jwilliams 18y agoThis is a slight tangent, but I think people would be surprised at just how many studies are based upon college/university students. I knew quite a few PhD students and their colleagues a while back - almost all of their test subjects came from the student body (in fact, in some places, students were had to participate in certain studies as part of their course requirements).
- robg 18y agoThis is true - but remember it's all relative. If most universities are drawing from approximately the same sample (e.g. university undergrads) then you essentially have a within subjects design. Still, that's why the science that survives is repeatable. As to the question of the MIT students, I recently saw a talk by this group and I have little doubt that the effects are an automatic function of visual memory. Participants just knew which objects they had seen. The results should replicate very nicely even at lower ranked schools. Indeed, I bet middle school kids would show the same effects. I don't think they recruited exclusively students though. I believe mixed in to the sample were folks off Craigslist. If you're really interested you could always send an email to one of the co-authors. You're more likely to hear back from one of the grad students.
- jwilliams 18y agoCouldn't help but think "MIT researchers find memory capacity much bigger than they remember" would have been a better title.
- whacked_new 18y agoAlso, it's visual long term memory.
- Eliezer 18y agoHow is this long-term? They were tested on recall the same day. Long-term memory would be a month later. Still, very impressive research - I presume the misunderstanding (or omission?) occurred at the press office.
- jsmcgd 18y agoI think the mechanism of short term memory works for about 30 seconds and memory recall after that length of time uses a different long term memory mechanism. So I think they are right to describe recalling events of 5 hrs ago as a function of long term memory.