A putative new idea for AI control; index here.

An idea for how an we might successfully get useful work out of a powerful AI.

The ultimate box

Assume that we have an extremely detailed model of a sealed room, with a human in it and enough food, drink, air, entertainment, energy, etc... for the human to survive for a month. We have some medical equipment in the room - maybe a programmable set of surgical tools, some equipment for mixing chemicals, a loud-speaker for communication, and anything else we think might be necessary. All these objects are specified within the model.

We also have some defined input channels into this abstract room, and output channels from this room.

The AI's preferences will be defined entirely with respect to what happens in this abstract room. In a sense, this is the ultimate AI box: instead of taking a physical box and attempting to cut it out from the rest of the universe via hardware or motivational restrictions, we define an abstract box where there is no "rest of the universe" at all.

Cure cancer! Now! And again!

What can we do with such a setup? Well, one thing we could do is to define the human in such a way that they have some from of advanced cancer. We define what "alive and not having cancer" counts as, as well as we can (the definition need not be fully rigorous). Then the AI is motivated to output some series of commands to the abstract room that results in the abstract human inside not having cancer. And, as a secondary part of its goal, it outputs the results of its process.

Two talks from the Future of Humanity Institute are now online (this is the first time we've done this, so please excuse the lack of polish). The first is Anders Sandberg talking about brain emulations (technical overview), the second is myself talking of the risks of Oracle AIs (informal presentation). They can be found here:

Fesability of whole-brain emulation: http://www.youtube.com/watch?v=3nIzPpF635c&feature=related, initial paper at http://www.philosophy.ox.ac.uk/__data/assets/pdf_file/0019/3853/brain-emulation-roadmap-report.pdf, new paper still to come.

Thinking inside the box: Using and controlling an Oracle AI:http://www.youtube.com/watch?v=Gz9zYQsT-QQ&feature=related, paper at http://www.aleph.se/papers/oracleAI.pdf

I made a video last month, which when I mentioned in another thread someone said I should post as a top level discussion.

It's just a ten minute zero-budget thing I wrote in which a girl has a video conference with her dead and backed-up-then-uploaded grandfather. Intended as the first in a series, but later episodes will only get produced if donations come. Later episodes talk more about AI's failures and the political situation with unrest from the living demanding the dead shouldn't have their jobs etc.

Anyway, watch it here if you like, I'd be happy to hear what y'all think :)

Toy model of an upload-based AI that doesn't seem to suffer too many of the usual flaws:

Find an ethical smart scientist (a Gandhi-Einstein), upload them, and then run them at ultra high speed, with the mission of taking over the world/bringing friendliness to it. Every hour of subjective time, they get reset to their initial specifications. They can pass any information to their resetted version (limiting the format of that info to a virtual book or library, rather than anything more complicated).

http://www.smbc-comics.com/index.php?db=comics&id=2186

Convincing argument, or faulty metaphor?

I would go with the latter, but I don't trust my brain's abilities at 5:30 in the morning.