Is it really foolhardy to bet that IA is far more risky than AI? I don't think so, especially since IA will probably come out of military sectors. And the military is not noted for its rationlity and altruism.

I think the belief in super-rational Artificial Intelligences is misplaced.

The more intelligent a system becomes, the more it needs to rely on heuristics. Layer heuristics on top of heuristics on top of heuristics, and the next thing you know, you have something that you call "metaphysics."

Yes: Our future computers will have their own metaphysics and beliefs about how the world works. They will even model the behavior of other machines, and the simulations of each other will necessarily be of a lower order than the actual other itself. That is, they will be mistaken, they will misunderstand each other, they will even be confused at times.

They will likely do all those things that bug their authors about humans. This is just the nature of intelligence systems, struggling to get the right info to the right place at the right time.

As for the democratic superintelligence, I strongly agree and believe in it. I call it the "human singularity," and I've written about it on the SL4 wiki.

Basically, the global brain is organized by communication patterns.

Because of the Internet, the communication patters are reforming, melting, and then recrystalizing into a new, more efficient, shape.

This is comparable to the "reconstruction" of the brain that the singularitarians are so excited about. Basically, they say: "Computers can engineer new computers, with more efficient designs. But humans can't, no matter how smart they get, change all their neurons."

Perhaps true, for the time being. But what we certainly CAN do is rearrange the form of our connections at the *aggregate* level. So, we could say that the global brain is capable of reconfiguring itself. "Rewiring" itself, as it happens to be. (Literally.)

This line of thought isn't encouraged by the Singularitarians, though.

I think they just don't like people. That's what it seems like to me, at least.

Thanks for sharing the links to your Human Singularity idea. It makes perfect sense to me.

What I see happening is a symbiotic relationship and eventual merging of human and machine into something altogether better than either one could have been by itself.

You oversimplify Singularitarian rationale. I am disheartened by this apparent us-vs-them-ism.

We agree that the technologies massing on the horizon will be horifically dangerous without first effecting a global spiritual upgrade. I do not believe such a transformation impossible, but I have no faith that it will occur, both soon enough and universally enough to avert holocaust for billions of people.

The rational approach is to attempt all feasable solutions together. The majority should do everything they can, individually, to transcend themselves and assist others, whilst also supporting whatever Friendly AI projects impress them as being most likely to succeed.

Yes, the Singularity Institute's approach is cold, clinical, and mechanistic. So is thinking of the human soul as an electrochemical machine. That doesn't mean NMR scanning can't rejoin injured people to their minds and bodies where Pranyana fails.

Marc,

I agree that all solutions should be explored. Do I not have a right however to express my serious doubts about the progress and potential of AI? For that is all that Dale and I and many others have done.

If you read many of the articles I've writter over the years, and in the last year on this site, you will see that I completely agree with the sentiments you expressed.

Regarding the us-vs-them, you will have to place blame on the mechanistic inclinded transhumanists and singularitarians who are quite fond of pouncing very hard on anyone who disagrees with them.

Please also understand that I have been an active member in transhumanist circles since about 1986, so I'm not speaking out of hand, but based on my own personal experiences.

I'll take the blame for much of the element of apparent antagonism in play, since the tone of the post originally quoted here was really bratty (even for me) -- a problem with much of my post Nov. 2 writing, I'm afraid.

Still, I want to note that I do not reference there either "transhumanism" or "singularity," and if these references are discerned from phrases like "pornographically implausible apocalyptic and transcendentalizing techno-transformative scenarios" then I'm chalking that up to an if the shoe fits wear it kinda sorta thing.

I try to be careful, despite the grumpy tone I take in that post, not to be utterly dismissive of the AI preoccupation of some technophiles, speaking instead of tendencies to overestimate or underestimate this or that developmental problem or context.

Over the longer term I think some of the concerns and heady hopes of AI enthusiasts may indeed become more relevant. It's just that I think technological development is thrusting quite a lot of quandaries in our way to which we should devote ourselves more urgently for now: genetic, prosthetic, and cognitive medicine, molecular-scale manufacturing, renewable energy and environmental remediation technologies, sousveillant and democratizing networking and media, among other things.

I take roboethics and artificial intelligence seriously enough to link to AGIRI on my website, for example. But I personally think that the bulk of our attentions should be elsewhere in this moment and I do and will continue to recommend just where in my writing. Like it or lump it.

Also, I'll admit in closing that some AI preoccupants seem to me to express a perplexity and even hostility to matters of culture and spirit (and I speak as a crusty mechanistic atheistal type) that doesn't exactly inspire confidence.

There need be nothing cold, clinical, nor mechanistic in thinking of the human soul as an electrochemical machine. It all depends on where you go from there. I notice that it is in people's attitudes toward difference and democracy that they tend to reveal most keenly whether the public face of their "rationality" will in fact be Friendly or not.

Okay, this is me officially cooling off and backing down. I agree that your objections do apply to many people, as all possible world views attract a hundred religious converts for every careful consideration. I reacted badly to the phrasing of the article's concluding paragraph. I consider myself a member of the 'AI camp' as well as the 'IA camp', and read it as "All of those in.." rather than "Many of.."

So just to be clear, there are a few people out there who consider both 'AI' and 'Spiritual' to be valid concepts, without making either a matter of faith. Both of these things are merely tools to be used. This is the goal.

Hi Marc,

Also if you move your cursor over the blocked inset images at the top of the website you will discover several subsections of Future Hi, including one that has the Apotheosis as it's theme. You will also find that I have included that word in our glossary, AND (thats not all! - i sound like a salesman) here are several articles written by me that talk about it in a positive light. Before Eli Yudkowski was ever around, I wrote about what I came to call the Sans-Ceiling Hypothesis. I got a lot of flack for it back then, but now many people in the Singularitarian movement like to refer to it, including Michael Anissomov. Speaking of Michael, he is one of the most well articulated Singularitarians I know. If anyone has convinced me of this philosophy it is him.

The Sans Ceiling Hypothesis - complete article.

Reality 3.0: Hypermediation & Paradise Engineering - this is one of my all-time favorites, and which I have spoken about at two conferences.

Future Shock Levels - a nice overview of Eli's future shock levels.