Topic: [ANN][MOTO] Motocoin - page 52. (Read 178273 times)

Yes, it only increases human mining margin by a very small (practically unnoticeable) amount.  In theory, if the anti-warp ever does kick on it would add more human margin since humans will generally do far less map iterations.

Once everything is sorted out with the difficulty warp, we can move straight onto implementing something like the "N heads" concept, or similar, which will do quite a bit more for the humans.

Another notion that I've been mulling over is the idea of allowing participants to "purchase extra lives" for use in their mining.  Basically, a user could burn some number of coins in a special transaction and receive an allocation of extra lives.  Then in proof checking, if the bike rider hits his head but a coinbase address for the block has some extra life the rider will not be put into a dead state, but the allocated lives will be decremented instead and the head would just "bounce" off of the ground.  I'm not sure if this would be utilized by people, considering that on the one hand rewind more or less accomplishes the same thing and only costs time, but on the other hand wall clock time for solution finding is in very short supply.  On the gripping hand, this approach might also just give even more advantage to bot operators, who could likely afford to buy a lot of extra lives and would probably do just about anything they can to increase solution space relative to search space!

It will probably take some combination of changes to make human mining feasible again, but we will accomplish it one way or another.  The only reason we haven't really been working toward this so far is the existence of the warp attack - priorities.

There are also some things that can be done "off chain" to help out the humans.  For example, I've been increasingly interested in exploring the idea of "cyborg mining" where a human and a bot work together to find solutions.  Bots easily get stuck wasting time in particular map layout scenarios where a human can trivially overcome the obstacles tripping the bot up.  Right now I handle this with an option to enable visualization on my bots, and can manually trigger a particular bot search thread to ditch the map it is currently "stuck" on when I see that it is stuck in one of these scenarios.  (This is the "metagame in the meantime" that I referred to previously.  Maybe I'll make a video of this in action to better illustrate what I'm talking about.)  When actively nudging my bots around like this I can increase their productivity by almost double when I'm intently focused.  I imagine that a more complete cyborg mining setup could detect when the bot was stuck, and present the game state to the operator to manually traverse the obstacle, who could then hit some key to pass the game state back to the bot to complete the rest of the proof.  Such a setup would likely blow both bots and humans out of the water, at least for awhile... probably until the third wave of bots start to come into the picture.