[notfed]

I still don't buy "then yes this is correct". If "ON" is defined as a state where ANY electromagnetic field flows into the lightbulb, then it is "ON" even before you throw the switch, because other photons are already shooting around all over the place, including from the neurons in your body pondering the result just before you throw the switch.

The definition of "ON" is exactly the main controversy with this and not a subtle detail.

[tsimionescu]

Even if we also accept the unspoken assumption that there are no other photons/electrical fields outside the experiment, the experiment is still wrong, since the battery will produce a teeny tiny electrical field even if the switch is off. This is much more visible if we assume the switch is at one end of the light-second long cables - throwing the switch has little impact on the experiment as explained.

[Ostrogodsky]

Yes but that does not generate clicks,views and that sweet youtube money.Sorry but I cannot stand those kind of guys (he, Mark Robber and the likes). Khan academy is doing something 1000000 times more valuable.

[kazinator]

"ON" is when the filament of the bulb is hot enough to emit far infra-red rays.