Forecasting framing renders neural emulation feasible

Before deriving an accurate model of how the brain works, we might be able to engineer neural emulations by simply framing the task as a neural activity forecasting problem. This reliably moves neural emulation in the “Thinking Humanly” quadrant, because the objective function essentially becomes “think as similar to humans as possible”, through neural activity as a proxy. However, the initial condition might also be crucial, besides getting the dynamics right.

Resources

Backlinks