Rant I wrote in IM

[ciphergoth]

I never post, so here's a rant written in IM I want to preserve. Edited somewhat

Comments

[bohemiancoast]

I think we might create improved intelligences and not notice that we've done so

[valkyriekaren]

Not sure I know enough about this to comment but it sounds like someone in this argument is either being dogmatic, or using terminology that is not clearly defined (i.e. not understood in the same way by both participants)

[seph_hazard]

I don't know if I'm a singularity critic or not, really. I have no idea how plausible it is, as I haven't done any research at all into what it would take and how likely we are to get there, but you seem convinced that we can and I trust your judgement and knowledge on these things. On a purely personal level I feel extremely uncomfortable about the idea, but I don't know how well-founded my reasons for feeling like that are. I reckon it's worth doing if it's possible, but I think I'd opt out.

[ciphergoth]

I'm not convinced that we will get there - but some people seem convinced that we will *not* get there, and that confidence seems unfounded to me.

[celandine13]

I hate those people too. The name-calling makes me angry.

I'm a skeptic because the history of the field of AI is strewn with the corpses of unsuccessful researchers. And since Turing computer scientists have been talking about intelligent machines. The whole 20th century has been a story of researchers thinking strong AI was around the corner. Why should today be any different?

[ciphergoth]

Hi, thanks for commenting!

So is there are particular way the Singularity could fail to happen that you're endorsing here?

It's been 56 years since the 1956 Dartmouth AI conference, which is as good a point as any to mark the start of efforts to build human-equivalent AI. We haven't succeeded in that time; how much does that tell us about the next 56 years? The next 112 years?

[celandine13]

My points of skepticism are somewhere between #2 (the idea of one mind being greatly more efficient than another might not be meaningful), #6 (... as above given the limitations of the height of our design ability), and #7 (We might not continue to study the problem, or the relevant fields necessary

[ciphergoth]

How likely does it seem to you that it will be forever beyond our abilities to scan and emulate the human brain, and run the emulation faster than the real thing runs?