Hi Eliezer,
I have some questions:
Could you tell us what you do on an average day (for example you were talking about making exercise) ?
How much time do you spend on thinking about AI each day and do you sometimes feel you are close to finding something important?
How would you define consciousness, and what mechanism makes it "emerge" in a mind ? (in other words : how can a network feel pain, for example ?)
Surely you don't have the answer to the second part of that last question, but how do you work to find it out ? Could you define your "thinking about thinking" process to resolve that problem?
Oh, and when will you post the damn videos of the singularity summit 2007 ?
Thanks a lot and keep on tsuyoku naritaiing.
Yesterday being my 100th Overcoming Bias post, it seems an opportune time to answer a commenter's question: Why am I posting?
For a long time I've suffered from writer's molasses. Like writer's block, only instead of not writing, I write very slooowly. At least when it comes to writing Documents - papers, book chapters, website material. If I haven't published a hundred papers, it's not for lack of a hundred ideas, but because writing one paper - at my current pace - takes four months full time. I sometimes wonder if I could become a respectable academic if I wrote at a respectable pace.
Oddly enough, I can write most emails around as fast as I type. Such disorders are hard to self-diagnose, but I suspect that part of the problem is that on Documents I repeatedly reread and tweak material I've already written, instead of writing new material. James Hogan (an SF author) once told me that he was more productive on a typewriter than a word processor, because the typewriter prevented him from tweaking until the second draft.
A blook is a collection of blog posts that have been edited into a book. Logically, then, publishing a book as a series of blog posts ought to be known as "blooking".
It would be more precise to say that I'm generating raw material to be edited into a book, and collecting some feedback along the way. I make no promises for this project. (I hate promising anything unless I have already done it.) The first part of the plan, generating the raw material as blog posts, has proceeded at a respectable pace so far. Will I be able to edit the posts into chapters, so long as all the raw material is there? Will I be able to generate all the raw material, or will the project, ahem, "blog down"?
In August I decided that I was going to write one blog post per day for Overcoming Bias. This challenge began to hone my writing speed somewhat - for example, I would look at the clock and try not to take longer than an hour... or three hours... but nonetheless I began to feel the need to shove the post out the door instead of perfecting it further. This is necessary and proper.
Near the end of August, I faced a new challenge - I also had to prepare two talks for the Singularity Summit 2007 (Sep 8-9). Those were actual Documents. I knew, from previous experience, that I couldn't possibly prepare the two talks and also keep up the pace of blogging on Overcoming Bias. Blogging was using up all my writing energy already - I have only a limited supply of words per day. If I overreach one day's budget I can't write at all the next day. So (I knew) I would have to temporarily stop blogging and resume after the Summit.
And then I said to myself, Hey, if I never try to do anything "impossible", I'll never grow.
I decided I would keep up the pace on Overcoming Bias while simultaneously writing my two Summit talks. Tsuyoku naritai!
I lost sleep, and skipped exercise. But I did it. I'll remember that the next time I'm thinking of trying something impossible.