lucid

Context: before going to bed I watched Trancendence. I was wandering whjat would happen if I try to build an AI, what would happen if the AI was able to reproduce, maybe a little mathmatical nonsense to this and the idea of all possible thoughts. Here is how it turned out:

Once in the dream I tried shaping a small unit, like a singele neuron, only difference: I tried to model it so big, that I could see it. I used a „thin tube“ as input and several larger tubes an an output, different thickness, being a higher prabability of conducting something like water poured into it. Concept seemed to work, constructs of larger size were able to solve „simple“ tasks like the TSP pretty well. Problem is, this modelling in RL would be a heck of work, slow, unpractical. Conclusion: I need to model more neural networks of different kinds, maybe one is RL applicable… well so much for this result of my dream.
Second I imagined a world where I supposed about every 6th part oft he soil was usable for constucting the robot I tried to set free. Only task: gather recources, reproduce. Given a slow circle to reproduce, the whole plain I was standing on was destroyed in a few minutes.
Memo to myelf: never ever let a robot build itself from corbon, oxigon or silicon… bad idea.
Next one I tried the Banach-Tarski-paradox. This enables you (in theory) to duplicate points in the 3rd dimension. It actually is quite fun to watch in a LD how it specifically works. Sad thing, this is not possible in RL…
Last one: the idea of the ending for Trancendence. I spend some time working out the details. The end is, (stop reading if you wanna avoid spoilers) yes, we can in the best case destroy an all powerful singularity, but maybe that’s not even the smartest thing to do. Sometimes external things know what’s best for us, better than we ever could. Like in politics, really:
We always cry, as soon as change hppens. Maybe we need to get a little more open minded.
With this I do not mean, we should accept everything, heck no, but we should try to see the good intentions, we should try to see the persons long term motivations, we should see what it has as consequences. For example, though this is arguably my personal püreference: we should see, that some things are unpleasent, but still must be brought through, taxes for example. Great medical applications like stem cells research. Some times you need to take a leap of faith. The world ahead can be a pleasent one… I think.