Waser's 3 Goals of Morality

mwaser

In the spirit of Asimov’s 3 Laws of Robotics

You should not be selfish
You should not be short-sighted or over-optimize
You should maximize the progress towards and fulfillment of all conscious and willed goals, both in terms of numbers and diversity equally, both yours and those of others equally

It is my contention that Yudkowsky’s CEV converges to the following 3 points:

I want what I want
I recognize my obligatorily gregarious nature; realize that ethics and improving the community is the community’s most rational path towards maximizing the progress towards and fulfillment of everyone’s goals; and realize that to be rational and effective the community should punish anyone who is not being ethical or improving the community (even if the punishment is “merely” withholding help and cooperation)
I shall, therefore, be ethical and improve the community in order to obtain assistance, prevent interference, and most effectively achieve my goals

I further contend that, if this CEV is translated to the 3 Goals above and implemented in a Yudkowskian Benevolent Goal Architecture (BGA), that the result would be a Friendly AI.

It should be noted that evolution and history say that cooperation and ethics are stable attractors while submitting to slavery (when you don’t have to) is not. This formulation expands Singer’s Circles of Morality as far as they’ll go and tries to eliminate irrational Us-Them distinctions based on anything other than optimizing goals for everyone — the same direction that humanity seems headed in and exactly where current SIAI proposals come up short.

Once again, cross-posted here on my blog (unlike my last article, I have no idea whether this will be karma'd out of existence or not ;-)

This is too confused to follow as a human, and much too confused to program an AI with.

Also ambiguity aside, (2) is just bad. I'm having trouble imagining a concrete interpretation of "don't over-optimize" that doesn't reduce to "fail to improve things that should be improved". And while short-sightedness is a problem for humans who have trouble modelling the future, I don't think AIs have that problem, and there are some interesting failure modes (of the destroys-humanity variety) that arise when an AI takes too much of a long view.

One of my frequent criticisms of LessWrong denizens is that they are very quick to say "This is too confused" when they should be saying "I don't understand and don't care to take the time to try to understand".

And if you can't imagine a concrete interpretation of "don't over-optimize" then you have obviously done no simulations work whatsoever. One of the most common problems (if not the most common) is to watch a simulation chugging along with all the parameters within normal ranges only to have everything suddenly rabbithole to extreme and unlikely values because of some minor detail (ruling out a virtually impossible edge-case) missing from the simulation.

Or, can you really not see someone over-optimizing their search for money at the expense of their happiness or the rest of their life.

Of course, this comment will rapidly be karma'd into oblivion and the echo chamber will continue.

That's over-optimising a single aspect resulting in overall under-optimisation.

It's not over-optimising overall.

12jimrandomh15y

The burden of clarity falls on the writer. Not all confusion is the writer's fault, but confused writing is a very major problem in philosophy. In fact, I would say it's more of a problem than falsehood is. There's no shame in being confused - almost everyone is, especially around complex topics like morality. But you can't expect to make novel contributions that are any good until you've untangled the usual confusions and understood the progress that's previously been made. If someone sacrifices happiness to seek money, the problem is not that they're doing too good a job of earning money, it's that they're optimizing the wrong thing entirely. An AI wouldn't see your advice against over-optimizing and put more resources into finding happiness for people; instead, it would waste some of its money to make sure it didn't have too much.

LESSWRONG
LW

LESSWRONG
LW

-16

Waser's 3 Goals of Morality

-16

-16