Bostrom

Nick Bostrom has written a book called Superintelligence, Paths and Dangers, Strategies. It is fascinating to read about the efforts that have been made to make AI. To be perfectly honest there is stuff he talks about I don't understand, and there's stuff there I could never understand so that is why he is an Oxford prof - Oxford is definitely the place I would go to find a white prof.

He began the book with a parable that talks of sparrows, owls and the sparrows inviting owls into their nests without knowing whether the owls eat sparrows. Except Scronkfinkle warns the sparrows, and he dedicates his book to Scronkfinkle. Whilst the sparrows of Scronkfinkle's nest might be dinner there would still be other sparrows. With AI the worst case scenario would be that there would be no sparrows left!! Am I being picky?

He begins his preface with "Inside your cranium is the thing that does the reading." Whilst I don't know of any humans without a cranium who can read, this statement makes me tirade, but not here ... maybe.

What is so fascinating is what they have been able to achieve. Years ago they were thinking that they couldn't invent a machine to win at chess, now it is done. In some ways this is impressive; I would get very little further than 4 moves against Gary Kasparov. But whilst the AI is beating me at chess, I had eaten a pizza, drank coffee, and watched a swan at the nearby lake. Meanwhile I was very grateful that another AI had cleaned the house ... and I am not going to mention Sophiabot who (which?) gave me pleasure in the bedroom this morning.

What I am getting at is that despite the great advances, the level of multi-tasking that women and some men can do, an AI cannot. So the question is whether I should have used "as yet" in that sentence.

Suppose we lived in a society where we are measured by our ability at chess, and only the best chess players survive, then what we have done is to invent a machine which will end our survival. Therefore before we invent the chess-AI we should have invented the HFP so that the chess-AI could not wipe out humanity. An obvious point.

Now there is a freaky but realistic scenario that we have to consider and that is I J Good's intelligence explosion - "their" terms as I don't see it as intelligence. We invent AI that can design AI, they design new AI that has more AI etc. There will be an explosion of AI that would make human intelligence appear minimal so why not swat the mosquito that is irritating?

This sounds a suitable doomsday scenario except for questions about intelligence.

Then there is developing intelligence from child's brains because it is considered easier to invent a child's brain, apply conditioning and experience, to get adult intelligence. And there is brain emulation. Get a brain - I can't remember whether Bostrom said it was dead or alive???, make this brain so that the AI has all the connections the human brain has.

But reading Bostrom is so infuriating. Sometimes I listen (in the car), and I have to turn it off - of the words I describe him the politest is fool. And this is an Oxford prof. There are two areas in which this occurs - spirituality and politics. Bostrom is working for British academia so indirectly he is working for the British government, NATO ... Trump-puppet. How responsible to the human race is it to deliver war-capable AI to the leading colonialists, to an alliance run by a country whose government dropped the bomb on Hiroshima. And Trump-puppet, MOAB ...

Where is the power? Without Oppenheimer there was no Hiroshima and Nagasaki. Without gunpowder where was British colonialism? Without scientists where would there be drone deaths and smart-bomb deaths? Scientists take this scourge off your shoulders.

Scientists need to stand up and put safety first.