Astonishing Edge on consciousness. Experiments show a clear, repeatable non-linear threshold between new events being in consciousness and not (the example is the masking effect of showing a subliminal word and then a string of letters; varying the gap between the two changes whether the subliminal is seen in consciousness or not).

The critical factor in whether these events are conscious or not appears to be whether disparate parts of the brain are active over certain time scales. Neural networks have been programmed to show similar behaviour.

One simple / simplistic thought is that these signatures are all there is. In that case, we could be said to have shown that we can simulate the existence of consciousness. However, it is not clear how these measurements of / reactions to novelty interact with our ongoing and continued perception of consciousness.

The suggestion in the discussion is that consciousness is an internal space for thought experiments, rather than an immediate unmediated reaction centre. The fact that there is a distinct in/out of consciousness also gives the system a digital aspect which is unexpected.

The experimental backup is fantastic:

  • Flash a subliminal number at you and ask you if it is greater or less than five and you perform at around 60%: better than chance
  • Flash one of four numbers and you get 40%: better than chance
  • Flash a number and ask for number + 2: you can do it
  • Flash a number and ask whether number + 2 is greater or less than five: you can’t do it

This suggests that the subliminal message isn’t in a processing system, only a recall one. Mind blowing.

And then the kicker: can this research into brain wave signatures help spot locked in patients? They have a test and are starting to apply it. Fascinating.

Next steps: decode brain processing, rather than simply detecting it.