Agree. One speaker at the workshop, Niccolo Negro, argued exactly this: in science, we usually try to avoid type 1 errors, but for C-tests, it's actually more critical to avoid type 2 errors (stating that there is no C when there is).
So workshop is to discuss current state and what to do about it.😔
Today I had an amazing opportunity to listen Philip talking about this study. Elegant design and cool analysis aiming at disentagling different neural contributions to natural scene viewing.👏
Interesting, I will take a closer look. Thank you! From the abstract, sounds a bit pessimistic though.
But the question of C-tests development is larger than AI. For example, if adult humans have P300 to consciously perceived oddball, does it mean that infants with P300 also consciously perceive...
Yes, from pragmatic POV, responsiveness markers are useful. But responsiveness may be present without consciousness, and one may argue, that with conscious beings, ethical considerations should be different, e.g. is it appropriate to boil lobsters alive if they are conscious and not just responsive?
The patient walks and stands upright,
But really nothing dubious;
The magic stimulation site
Is subthalamic nucleus.
#brainrhymes