The weather today here is terrible, following what was for me, as a bicycle commuter, a tiring week. After several weeks of riding very little while on vacation, I came back and worked a full week (imagine that!) and rode the "Polar Vortex" on Tuesday and yesterday in freezing rain in the morning when I managed to find the one icy spot between my house and work and fall down. (I did not suffer much physically from the fall, it was just unpleasant and inspires further thought about how to avoid such things in the future.)
Anyway, I happened to bump into this video on YouTube that I find very entertaining - yes, it is several years old. It only has around a million views, however, so not everyone has seen it . . .
Robot manages to keep the fixed gear bike upright with internal controls
The general direction the robot takes and the speed of travel is controlled by someone with and RC unit but keeping the bicycle upright during travel is up to the robot itself as an independent activity. The robot rides a fixed (single) gear bicycle that simplifies things somewhat - the robot doesn't use a braking system to slow the bike down but simply pedals more slowly until it is almost stopped, then puts its feet down quickly to keep the bike upright.
The person who created this thing states, "I'm interested in artificial intelligence, and in that context I think intelligence and skills have equal value." I am not even vaguely an engineer, but I am someone who rides a bicycle a lot and there is sometimes (rarely, but still) a conflict between the direction I want to have the bicycle follow and the almost reflexive "keep the bike upright" moves my hands make with the handlebars - it can be a little unnerving. Here, the operator steers the bicycle only approximately since the robot must make independent adjustments with the handlebars to keep the bicycle balanced. To me, it isn't whether the intelligence to choose the route and the skills to keep the bicycle upright are equal, it is that they are independent and if not managed intelligently can be in conflict - for example, with poor planning on my part, I end up running off the edge of the trail because my reaction at some point causes me to steer off the trail to stay upright because I choose (with my intelligence, or lack of it) to ride to close to the edge otherwise.
This kind of thing seems particularly relevant for "intelligent" drone robots that will be flying overhead, it seems. They will also combine intelligence to choose routes with reactive flying skills that may (or may not) have the same issues.