[petermcneeley]

I have always assumed that if I explain something to someone and they dont get it then I need to do a better job at explaining it. Its like the difference between a magician and a scientist.

A Magician takes the trivial and shrouds it in confusion till one is astonished by the reveal. A scientist deconstructs the reveal into the simple and trivial.

I loath the magician.

[bachbach]

Well, there is simple and correct and simple and wrong.

The Copernican description of reality is simple and correct and threw away complex but incomplete models.

I'm sure you've noticed that people frequently create complex models for something, but it is only later it gets reduced to something easier to understand.

Probably this is because the elegant model is difficult to reach on the first pass, so the more circuitousness paths are taken first - your thoughts?

[petermcneeley]

Interesting, have you read this? https://chem.tufts.edu/answersinscience/relativityofwrong.ht...

[bachbach]

Yes, it's interesting.

Not appreciating the mental scaffolding that takes you to the next point is a serious problem. Developmental economics and artificial intelligence - the scaffolding is neglected.

In developmental economics we only thought we knew how we got rich. Regressive planning (working backwards) failed because we are unaware of most of our past, even the recent past - we had a false consciousness about what actually took place. What we have is a bunch of just-so stories for grown ups and this is normal for humans in everything.

Neuroscientists say this is feature of how our memory works.

AI is similar because of Moravec's Paradox. People have been going around saying that more intelligent machines will come and replace physical labour for some time but the opposite is true. The paradoxical results of Moravec's Paradox occur because again we are unaware of the scaffolding required to replicate an artificial cerebellum system aka a brain/nerves/muscles - a body that knows how to move.

If you watch the most recent videos of walking computers you'll notice we are heavily implying we need very effective computer algorithms and vast quantities of data to produce a robot that can walk across the floor.

There is something wrong with that picture. I wonder if sea slugs brute forced traveling across the ocean floor with their supercomputers (now lost to us!).

[petermcneeley]

A slug doesnt need a supercomputer because a slug is working more like this thing https://www.youtube.com/watch?v=qv6UVOQ0F44