Surprisingly Good Solutions?

Tom McCabe writes:

A superintelligently-designed Utopia should contain a large number of qualia which are surprisingly good, in the sense that we’ve never even imagined that point in utilityspace. Such qualia are relatively rare in the present world, but they do exist; imagine a starving African child eating their first chocolate cake. This, in some sense, is a more worthy goal for the future than, say, Iain Banks’s Culture series. The Culture is limited by human imagination, while superintelligently-designed utopias shouldn’t be.

Do we agree with him?

I wasn't convinced on Overcoming Bias a few weeks ago:

It's also interesting to wonder whether the goals of "radical" transhumanists might be a little self-contradictory. With a limited human brain, you can (as a matter of physical fact) only entertain thoughts that constrain the future to a limited degree. Even with all technological obstacles out of the way, our imaginations might place a hard limit on how good a future we can try to build for ourselves. Anyone who tries to exceed this limit will end up (somehow) absorbing noise from their environment and incorporating it into their preferences. Not that I have anything against this - it is how we got our preferences in the first place - though it is not a strong motivator for me to fantasize about spending eternity fulfilling preferences that I don't have yet and which I will generate at random at some point in the future when I realize that my extant preferences have "run out of juice".

This, I fear, is a serious torpedo in the side of the transhumanist ideal. I eagerly await somebody proving me wrong here...

But now I'm not so sure. I invite readers to pitch in.

Responding to Anne Corwin on "dreams of immortality"

Over at Existence is Wonderful, Anne Corwin has a post up called:

Thanks, But I'll Take Test Tubes of Enzymatic Slime over 'Fountains of Youth' Any Day

Now, this title and the ensuing post annoyed me, because I think it is exactly the wrong attitude to take towards life extension research. I think that Anne's position is an intelligent, even semi-coherent position, but I think it is a bad approach to take, and I hope that by critiquing it I can add a valuable sense of direction to the h+ and life-extension movements. First things first: let me paraphrase Anne's post. In fact, I'll just quote her, leaving out things that she says which I don't disagree with or don't want to comment on here. Readers are encouraged to read the original post and correct me if I have significantly altered Anne's intended meaning:

... I am very sensitive to what my rhetoritician friend Dale Carrico calls superlativity.

... I didn't initially see what was wrong with going on about "technological immortality". But now I do.

... I do not claim that any of the developments being discussed are somehow representative of an inevitable convergence of technology toward humans being able to run around forever in indestructible robot bodies ... That sort of thing is fine if you're writing science fiction stories, or even just daydreaming for the fun of it, but it has nothing to do with what can be accomplished within a reasonable timeframe given what humans know now.

... Superlative predictions tend to tempt people toward glossing over the "...and then a miracle happens!" step between now, when we still have people dying at 90 of heart failure and a potential future in which five-hundred-year-olds run weekly marathons (or accomplish some other feat associated culturally with vigor and youthfulness).

... I certainly don't think such a future is impossible, but neither do I think one needs to believe in such an outcome in order to make healthcare improvements (for the elderly and everyone else) a priority.

... There is no magical property to handwaving that somehow makes handwaved potential outcomes more likely to happen.

... [Longevity research at ASU being described as researchers "seeking a fountain of youth"] - might be attention-grabbing, but at what cost? Do we want to invoke legends and myths in describing interesting and potentially promising science experiments, or the reality of present-day healthcare needs (and potential real-life improvements thereof)? I would definitely lean toward the latter. It may be less glamorous, but when it comes to science and ethics, I'll take the hard (but ultimately more promising) practicalities of the present reality over daydream-invoking glamour any day.

The debate started in the comments. My second comment contained this:

if people don't have hopes and dreams to motivate them to do/fund the research, they just won't bother.

To which Anne responded, including:

You seem to be insisting (and please correct me if this is a misinterpretation) that people who see it as problematic to talk about longevity in terms of "immortality" and "fountains of youth" are somehow the ones harming the prospects of research.

That doesn't make sense.

Daydreaming and the science fiction can certainly enrich and inspire people's imaginations -- heck, the reason I'm an engineer now has a lot to do with my enjoyment of science fiction growing up.

But to suggest (by analogy) that in order to get to the moon it is somehow necessary to believe in (or sit there hoping really, really hard for) transporters and warp speed is silly.

Being inspired by something and having motivation is not the same as acting like the images that inspire you must be "defended" as potential realities.

That last sentence is telling: Being inspired by something and having motivation is not the same as acting like the images that inspire you must be "defended" as potential realities. Come again?

Ok, let's get down to basics. A poll for my readers: would you spend time, money and effort on anti-aging research if you did NOT think that an extended human health span was a potential reality at some point in the future? It's a no-brainer. Either you think that extended human healthspan is a realistic possibility or not. If you do not think that it is a realistic possibility, then why are you spending time and effort trying to achieve it? And yes, I think that you must defend that outcome as a potential reality; if you can't defend it as a potential outcome, why are you working on it yourself?

Now we come on to the most important disagreement, and one where I hope to change Anne's mind. Both Anne and I agree that funding longevity research today could result in human physical immortality. The mechanism goes something like this:

If we can extend your life enough that you're alive in the year 2150, say, then the technology available then may be rather a lot better than it is now, allowing your lifespan to be extended further. By this point, we are very uncertain about what the world will be like. Old certainties like death might not be around. There are many mechanisms by which death could be abolished in, say, the 22nd century, including mind uploading, progress in AGI, progress in nanotechnology, etc. We can see some scenarios where people would not die. We can see others where they would die. The specific dates mentioned here are not that important. We can think of scenarios where mind uploading happens this century. We can see scenarios where life extension research extends people's lives by more than 100 years, etc, etc.

This argument suggests that immortality is possible, and that it is a non-negligible possibility, and that for some group of people (a rather large group, remember 100,000 people die of old age PER DAY), avoidance of death may depend upon whether or not anti-aging treatments are researched today.

So, should we tell people this? Or, should we de-emphasize this particular aspect of the truth and instead say something like:

"Breaking news: Scientists are working to explore the use of microbial machinery to biodegrade 7-ketocholesterol"

People respond to hopes and dreams, not to oxidative pathways and tiresome details of microbial machinery. If you tell them that a certain line of research may allow them to live forever, and they believe you, they will respond by giving you lots and lots of money to make sure that the research gets done. If you tell them that the exact same research is "exploring biodegredation of 7-ketocholesterol", they won't even hear you.

Personally, I think that it is unethical not to tell people about the potential for immortality and the research that will get us there, in the same way that it would be unethical to "forget" to tell someone about the existence of a chemical that might be a cure to a poison they had ingested. These situations are ethically isomorphic.

Yes, Anne, the possibility of superlative outcomes "distorts" people's thinking, in the same way that if you had been bitten by a snake and I told you that there might be an anti-venom in the locked medical cabinet, your thinking would be "distorted" into doing everything you possibly could to get that anti-venom. This is a good thing.

But I feel that this argument will not sway Anne, who will insist that these are totally different situations (perhaps she will claim that, for some unknown reason, the best way to get the anti-venom to the snake bite is to pretend it doesn't exist? After all, it might not work, might not be in there, etc).

So, as a last ditch effort to convince Anne that superlative dreams are a good thing, I point her to a common enemy:

God's purpose for you

If people don't see a rational way of bettering their existence, they will clutch at other "solutions":

The purpose of humanity is to become the spiritual children of God! The purpose for your life is more than making money, gaining social status and wearing the right clothes. The coupling of the human spirit with the Spirit of God makes possible the development of a new nature and eventually a resurrection to a new life as immortal children of God, joint heirs with Jesus of all things.

Viable psychological niches that are compatible with doing something about death?

Bostrom, De Grey and Savulescu declare war on death

My angle on it:

One question to consider is this: why is anti-aging such a hard idea to sell? Many people who I have broached this topic with either:

(a) are totally apathetic and/or quickly change the subject

or

(b) find some obscure argument that they claim is totally compelling as to why we shouldn't do this. I have heard the usual ones: death gives life meaning, we must kill ourselves to make way for future generations, we mustn't do anything about the problem until every last African is well fed, death is a great way to get rid of dictators (and these people will claim that dictators are bad because they kill a few thousand people a year), etc, etc.

or

(c) will argue at great length that curing human aging and death is *obviously* impossible and that I need to come to terms with my mortality, even in the face of fairly strong arguments to the contrary and even when I know much more about the relevant science than they do. People sometimes make up arguments on the spot as to why SENS and/or cryonics won't work.

There is probably an underlying psychological explanation for this behavior. (a), (b) and (c) probably correlate with three distinct ways of coping with our mortality:

(a)-type people , who are usually young, can just pretend it doesn't exist. There's no way in hell they'll get involved with the anti-aging movement, because then they would no longer be able to pretend that death doesn't exist.

(b)-type people, typically religious or environmentalist or very left-wing, have decided that death is a good thing after all, and they are frightened that their protective worldview will collapse if the anti-aging movement succeeds. These people may actually sabotage the movement, bomb rejuvenation clinics, etc. Imagine all of these youthful, wise, rich, popular, happy, successful 100 year olds swanning around in the year 2050 as living proof that death is actually not a good thing. That would be anathema to the (b) worldview.

(c)-type people, usually materialist rationalists, can cope with mortality, but only if it is TOTALLY unavoidable and thus not to be worried about. The idea that aging might be cured at some point in the future is poison to their particular psychological defense: it is only by living with the absolute certainty that death is unavoidable that they can relax about it.

I think that the most potent weapon in the war on death right now is to examine the psychology of those who implicitly support it, and find ways of giving people viable psychological niches that are compatible with doing something about death. This will involve talking to some psychologists and perhaps doing some research.

Minsky, Sloman on commonsense intelligence architectures

Marvin Minsky and Aaron Sloman are experts on AI - they've spent their lives doing it. Their best guess for an architecture for an AI with commonsense or "human level" abilities is the following:



Mental spaghetti, anyone?

"Minsky describes “ways-to-think”
as configurations of agents within the mind
that dispose it towards using certain styles of
representation, collections of commonsense
knowledge, strategies for reasoning, types of
goals and preferences, memories of past expe-
riences, manners of reflections, and all the oth-
er aspects that go into a particular “cognitive
style.” One source of knowledge relating prob-
lem-types to ways-to-think is the causal diver-
sity matrix discussed at the start of the meet-
ing—for example, if the system were presented
with a social problem, it might use the causal
diversity matrix to then select a case-based
style of reasoning, and a particular database of
social reasoning episodes to use with it.
However, any particular such approach is
likely to fail in various ways. Then if certain
“critic” agents notice specific ways in which
that approach has failed, they either suggest
strategies to adapt that approach, or suggest
alternative ways-to-think, as suggested shown
in figure 6. This is not done by employing any
simple strategy for reflection and repair, but
rather by using large arrays of higher level
knowledge about where each way-to-think has
advantages and disadvantages, and how to
adapt them to new contexts."

The researchers present clearly didn't have much of a clue about what the use of a human level AI would be, though:

After a good deal of argument, several partic-
ipants converged upon a vision from The Dia-
mond Age, a novel by Neil Stephenson. That
novel envisioned an “intelligent book”—The
Young Ladies Illustrated Primer—that, when giv-
en to a young girl, would immediately bond
with her and come to understand her so well as
to become a powerful personal tutor and men-
tor.
This suggested that we could try to build a
personalized teaching machine that would adapt
itself to someone’s particular circumstances,
difficulties, and needs.

Passages like this remind me that the field of the future of humanity is a weird place to be. Minsky, Lenat, Sloman, and whole host of AI experts gather at a conference to discuss how to build the most powerful machine in the universe. In my very humble opinion, their ideas are fairly good. Then they decide that they need a use for it - and decide that personal tutors are the most compelling application of a human level AGI!

Read the original article here.

XKCD on Artificial Intelligence Risks

See the original here