4 ms·
> It's easy to show crabs can make logic gates - just do it That's just an artifact of the way we reason about the world. We take as an assumption that inducti
by hansvm 6y ago
> It's easy to show crabs can make logic gates - just do it
That's just an artifact of the way we reason about the world. We take as an assumption that induction works for "positive" events (we showed that these crabs can make logic gates) and can't accept the same evidence when it pertains to "negative" events (we showed that crab movements _are not_ uniformly random -- no you didn't; what if you didn't feed the crabs the right kind of peanut butter?).
- Dylan16807 6y agoIt's easy to prove "X is not always true". It's hard to prove "X is not ever true". Your "random movement" hypothesis is the former, so this evidence would in fact be accepted. The wombat hypothesis is the latter.
- hansvm 6y agoYes, but the hypothesis in the paper (soldier crabs can implement logic gates) is _also_ of the latter form, despite its positive construction. Being usable for a logic gate implies an element of control and repeatability, and unlike a statement like "the sky can be blue" a point observation doesn't suffice as proof.
- Dylan16807 6y agoI disagree. A moderately unreliable logic gate is still a logic gate, and no reasonable person would say that the crabs are guaranteed to work. Look at the actual paper. A full third of it is about the robustness of the gate.