Shortly before leaving MIT AI Lab, Rodney Brooks gave a talk in which he pointed out we can’t build something as intelligent as a bacterium.
Hmm. I worked in this area 20 years ago. At that time, it wasn’t capable of predicting biological properties of 10-atom systems
-
-
…which was because of inaccuracy in the underlying physical model, not computational limitations.
-
People ran simulations anyway, because they were desperate for any insight, and it could provide some qualitatively sometimes
- 3 more replies
New conversation -
-
-
There are still the same tradeoffs: model fidelity vs simulation speed, and timestep size vs simulation duration.
-
You can't do the attosecond quantum stuff and the multi-second signaling pathway in the same model. Still need hierarchy and abstraction.
End of conversation
New conversation -
-
-
Though our number mostly line up - 2 decades, 4 orders of magnitude growth in CPU ~-> 10^5 atoms at about the same fidelity.
-
At the other end, eg planetary atmospheres, tiny input uncertainties lead to huge changes in what goo gets formed over billions of years.
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.