A coherent and actionable goal for AI?

Can the following goal proposed for AI be successfully implemented?

“Our coherent extrapolated volition is our wish if we knew more, thought faster, were more the people we wished we were, had grown up farther together; where the extrapolation converges rather than diverges, where our wishes cohere rather than interfere; extrapolated as we wish that extrapolated, interpreted as we wish that interpreted.”

http://intelligence.org/files/CEV.pdf

On Suicide

The thing about saying anything less than negative about suicide is that your audience will try to blame the position on the fact you’re feeling suicidal, a position they are certain you would not hold if things were going well.

So the best time for me to write on this topic is now, at a time that I’m feeling quite positive.

Suicide. I tend to believe that a good number of suicides, perhaps a majority, were attempted while the individual was short-sighted. Life most certainly has a way of evolving out of a bad spot into something quite amazing.

I was suicidal after my divorce 2 decades ago. A depression held me in its grip for more than 2 years. The pain of the divorce kept me in a dark, but very introspective4 and creative mode. As I slowly emerged from the darkness and looked around me, I realized that life can morph into hundreds of different and brighter forms. I’m so very glad I did not end my life at that time.

However, this is not always possible for everyone at every time. There are situations that allow little hope for future happiness. This is seldom the end of a relationship. It more often occurs when a sickness or disease makes life unbearable due to the physical or emotional pain.

Age is also a factor. Each year I feel more and more satisfied I’ve lived life. There are still many things I intend to do, but I don’t think anyone can say Phil did not live a very full life. So, if I found myself experiencing excruciating pain with no hope of relief, I’d have no problem considering suicide as an option.

While acute and relentless physical pain may be a common legitimate reason to contemplate suicide, I think there are very few cases in which emotional pain would warrant suicide. Admittedly I base this on the fact that I came out of, what I considered to be, an extremely painful depression after my divorce. There may be degrees of depression I have not experienced, or situations in which the emotional pain does not subside. But I tend to doubt it. I don’t think I’ll ever again think about suicide simply due to emotional pain.

The following are the factors I think everyone suffering physical pain should carefully weigh before considering suicide.

The severity of your pain.
The hope of recovery.
Your potential productivity in spite of the pain, giving special consideration to the degree that you can help others.
The degree to which you have already had a great life.
Those negatively affected by your death.

The factors you should not consider include the following.

What others think is the “moral” decision for you to make.
The way your death will be a way to “get back at” or “take revenge on” anyone.

I would suggest that, in most cases, consulting those you love is a good idea. And you definitely do not want to leave anyone you love uninformed about your reasons, leading to them blaming themselves in some way. But ultimately, it is your life to end. If you have family members who would rather see you in immense pain than to have you missing from their overly-dependent lives, don’t feel guilty about deciding what is right for you…and probably also best for them.

Well, that was a bit too somber. Time to get back to living!

Good News for Apologists

Breaking News

In what Biblical scholars are calling the greatest doctrinal breakthrough in over three millennium, close associates of Jehovah have revealed that, what started as a simple prank, led to theologians writing hundreds of millions of pages of unnecessary convoluted arguments. It was Thor who, in an exclusive interview with Larry King, revealed that he, along with Aphrodite, managed to gain access to Jehovah’s iPhone while the Almighty was expressing to William Lane Craig the immense anguish braved by the Israelite soldiers whom He commanded to slaughter infants.

The divine pranksters altered the auto-correct feature on Jehovah’s iPhone to have some words automatically replaced with their opposites at the dawn of creation, leading to a hilarious three millennia of awkward apologetics. The inverted terms included the following.

“Hate” was changed to “Love”.
“Unjust” was changed to “Just”.
“Impatient” was changed to “Patient”.

The revelation of these changes definitively resolves the bewilderment theologians perennially experienced while attempting to reconcile the behavior of Jehovah with claims about his character. Thor and Aphrodite giggled for ages while apologists anguished over the notion of a god who becomes so wrathful over a single offense that he can think of nothing better to do than to damn to Hell the offender, and yet who is called “loving”, “just” and “patient”. The red-faced theologians now admit the revelation makes obvious sense, and have admonished the faithful to accept this far more sensible description of Jehovah’s character that is indisputably reflected in His 1) commandment to slaughter infants, 2) refusal to condemn slavery, 3) requirement for rape victims to marry their rapists, and, of course, 4) damnation to Hell of any sinner who follows a sin nature they neither requested nor can avoid.

Snorting his bemusement, Thor revealed to Larry King other, perhaps less disruptive, inverted terms.

“Speak to the finger” was changed to “Pray”.
“Screw Africa” was changed to “Amen”.
“Marvin” was changed to “Lucifer”.

When asked why they would do such a thing, Aphrodite invoked the long history of Jehovah co-opting pagan holidays for His own worship. Other deities are reported to have found the prank over-the-top as it has led to a prolific volume of painfully perverse YouTube videos defending this impossible description of a loving, just and patient Jehovah acting hateful, unjust, and not just a little impatient.

Jehovah refused calls on the incident, but it is rumored that Thor’s hammer was spotted hanging high on a high-school flagpole along with Aphrodite’s most recent Victoria’s Secret purchase.

Elaborating on Induction and the Inappropriateness of Presuppositions

{The following is a post I made on a thread on the Facebook group Unbelievable.}

As I mentioned earlier, induction is assessing what works. I’m not merely claiming induction include assessing what works, it IS assessing what works. And rationality is positioning one’s degree of belief to a degree that maps to the degree that something has been assessed to work. 
So you have the following apparent paradox: what would it mean for the assessment of what works to fail an assessment that it works? :) This is nonsensical.
Let’s apply some rigor and tease out some of these concepts.
What do we mean by “what works”. This is essentially noting regularities in our experience. These regularities allow us assess to what degree we are warranted in believing event X will occur in the future given its regularity in the past.
And there is what we might call a meta-regularity: the observation of regularity in the success of assessing regularities. Herein lies the apparent paradox: Is it possible for our assessment of regularities become irregular? As fallible minds we can not dismiss this possibility even though we can not fathom what this could ever mean. If this assessment of regularities ever becomes irregular, would there be something else that provide us with the power of predictability? In this remote scenario, I would simple back away from my confidence in this inductive process, and explore whether there was something else that would make my reality more predictable. We are warranted in following whatever works to make our realities predictable. We place non-absolute confidence in what appears to work to the degree that it appears to work, and for as long as it appears to work. No presumptions are required.
(This is parenthetical to some degree, but if I were to discover that I was a brain in a vat, I would not protest, but simply think “Well, this is interesting.”  It appears others who have decided they are somehow warranted in absolute certainty in logic, mathematics or other axiomatic notions would feel the need to protest the violation of the impossibility of this new revealed reality…or so it seems from some of the comments claiming some absolute presumptions are warranted.)
Now let’s deal with the notion of rationality. I want to start with the concept and then find an appropriate linguistic tag rather than start with a linguistic tag and subsequently try to force a concept into that tag. The concept is following what works, and includes the connotation of value or praise being placed on following what works. This is made most salient with its converse. We all consider following what we perceive to not work to be irrational. (This is a critical point, so if you disagree, make that clear.) This is a gradient concept, so we can add nuance by restating this as, “to the degree that someone knowingly follows what they perceive to not work, to that degree they are irrational”. The converse of this produces what I, for a couple decades, have been calling “rationality”. The inverse correlative would be “to the degree that someone knowingly follows what they perceive to work, to that degree they are rational”.
Now, this is important. For this concept, I’ve assigned the linguistic tag “rationality”. Words belong to convention and are not bound by my stipulations, but it appears that most conventional uses of the word “rationality” at least intersect with this concept. If you are in disagreement and have a better linguistic term that will better convey the concept, please let me know. If you feel that the term “rationality” does not adequately convey this concept, provide a term that will more effectively convey the concept. It is the concept I want to convey rather than promote a linguistic tag since concepts are logically prior to the terms employed to convey those concepts.
It is essential to this project that we use terms for which we share identical referents. I’m going to assume you are, at this point, with me on this definition of “rationality”.
Now, note that throughout this discussion I have not treated belief as binary. Nor should I. Because rational belief (in propositions other than immediate perception) is a degree of belief commensurate to the perceive balance of evidence, and because evidence is inductively assessed, all rational beliefs are sub-absolute and subject to revision. There is no need to carry presuppositions, nor is there warrant for this. If by “presupposition” you actually mean a sub-absolute degree of belief wholly based on the degree of the perceived evidence, make that clear now. I would suggest that the etymology of the term “presupposition” implies it is the action of irrationally accepting without any degree of doubt something prior to its assessment, and suggest that it would be confusing to suggest a presupposition is a rigorous mapping of the degree of belief to the degree of the evidence.
(When I say evidence, I don’t only mean only empirical evidence, but also the evidence provided by logical assessments of the proposition in question.)
With this foundation, I think I’m now ready to introduce an argument why you are not rational in holding any presupposition.
P1: For any presupposition X, there was a time in your existence when you did not believe X.
P2: If there was a time where you did not believe X, there was necessarily a point at which you mentally acquired a belief in X.
P3: Human mental acquisition is fallible.
P4: Anything mentally acquired by a human may be wrong.
C1: Your human presupposition X may be wrong.
P5: All presuppositions are believed with absolute certainty.
P6: Presupposition X is held with absolute certainty.
C2: Presupposition X may be wrong, yet is believed with absolute certainty.
P7: Whenever the degree of belief in a presupposition does not map to the degree of perceived evidence for a presupposition, that presupposition is held irrationally.
C3: Presupposition X is held irrationally.
Here’s a more pictorial description of this logical relation. Deduction is subsumed within induction. Deduction necessarily lies within the logical space induction, and can never, for fallible human minds, violate that constraint. All the axiomatic foundations of our deductive games and tools have been acquired at some point in our lives through inductive processes.
Now for the reiteration of my conclusion. I hold no presuppositions. I know my degree of certainty for everything I believe, and there is nothing that I hold with absolute certainty. (I think we agree that it would be absurd for person A to inform person B that they are mistaken about the degree to which they believe something.) Not only do I hold no presuppositions, I have argued that, holding presuppositions is intrinsically irrational.
Let me introduce one more salient example that may clarify some things.
Imagine you are being chased by a bear, and you come to an old suspended walk-bridge across a valley that would allow you to escape…if the walk-bridge does not snap. You assess the frayed ropes of the walk-bridge, and conclude that there is perhaps a 50% chance that you’ll make it across the walk-bridge. You assess there is a 1% chance you’ll survive if you turn to fight the bear. So you decide to cross the bridge. This choice is binary. Either you attempt to cross the walk-bridge or you don’t. You have chosen to attempt to cross the walk-bridge.
Now, does this decision to attempt to cross the walk-bridge require you to change your assessment that the walk-bridge has a 50% chance of snapping? Of course not. If you did, you would be irrational. Yet, I have actually heard it suggested, often in the context of a religious proposition, that it is both rational an necessary to abandon your honest assessment, and to believe the walk-bridge will not snap (or whatever the proposition is) with absolute certainty. This is wrong. Rationality is mapping our degree of belief in X to the degree of the perceived evidence for X. And decisions don’t affect nor should affect that honest assessment of your proper degree of belief.
===============
{The following is an earlier post, and might add some clarity.}
 The crux of your misunderstanding is found in your notion that induction may not be reliable.
What would it mean for induction to suddenly become unreliable?
Induction is simply identifying regularities.  It doesn’t just involve that, but is equivalent to that. What could it ever mean to identify irregularities in a project of identifying regularities? What could it ever mean, in other words, to find induction irregular?
If there was something that worked better than induction to make sense of the world around me I would abandon induction for whatever is more reliable. (I hope you get the joke here.) (Also note that induction is not cause-and-effect. Readers of Hume may be confused in this respect.)
Rationality is following what appears regular to the degree that it appears regular and for the duration it appears regular. There is no commitment to the persistence of the regularity.
My confidence in any notion about the world around me is grounded in its regularity as I perceive it. When that falters, my confidence falters. When it rises, my confidence rises. I need not absolutely believe any axiom. Axioms are useful only within a system that you, if you are to be rational and epistemically humble as a fallible mind, have not excepted with absolute certainty.
So, no. I’m not required to presuppose anything. Nor do I. This is what epistemic humility requires.
(Note that the type of presuming/presupposing I’m rejecting is one of absolute certainty and does not include ‘assuming for the sake of argument’ or ‘for the sake of testing’ scenarios.)
=====
Comments are much appreciated.

The Irrational Worm in the Rational Apple

I’ve been promoting rationality on the web for some time now. There is a good core rationalist moment that sticks to the basics of identifying and eliminating logical fallacies, cognitive biases and other barriers to clear thinking. However, many groups of skeptics and rationalists have been co-opted by those with some agenda that they are promoting with very irrational methods. Let me make a few points.

You are not rational if you…

  1. consider arguments contrary to your opinion to be insults
  2. respond to arguments with insults or sarcasm rather than with counter-arguments
  3. think participants in a discussion must be of a particular race, sex or ideology instead of simply being capable of presenting arguments
  4. create a false dichotomy by suggesting that any rational assessment of the strength/weaknesses of accusations of X is actually a defense of X.
  5. tell others what you have concluded they believe instead of asking them what they believe
  6. intentionally misrepresent someone’s position rather than interpreting their comments graciously

There is currently far too much irrational nonsense emitted from those claiming to be rational.

I’m calling it out when I see it from here on.

Rationality is Prior to Burden of Proof

Epistemic rationality is honestly positioning one’s degree of belief in a proposition to the degree that the balance of evidence for that proposition warrants. Epistemic rationality belongs to the individual epistemic agent.

Burden of proof is more of a social concept. If someone wants to convince members of a community a proposition that is incongruent with the current beliefs of that community, then, to accomplish that goal, he or she needs to take on the onus of presenting the arguments/evidence that will accomplish that.

Note that this assumes that the community values arguments and evidence. What could it mean to have a burden of proof where proof (read “convincing evidence” here), is not valued among epistemic agents. If the epistemic agents within a community hold that beliefs need not map to the balance of evidence, then the notion of burden of proof will be of no value to them.

So, what is logically prior? Epistemic rationality is prior to burden of proof. If a community thinks, to any degree, that they need not map their degree of belief in X to the degree that X is warranted by the confirming/disconfirming evidence, then they only foolishly inform those who would attempt to provide arguments/evidence for a position not their irrational own that they must take on the burden of proof.

Burden of proof is like the status of offense or defense in a football game. If you don’t respect the rules of football, or want to play football under the rules of tennis, you don’t have any right to suggest the other team must now play offense. Until you demonstrate a willingness to play by the rules of football, you don’t have any right to suggest who takes possession of the ball.

Rationality comes first. Only after you have committed to mapping your degree of belief to the degree of the confirming/disconfirming evidence can you suggest the other side has the burden of providing evidence/argumentation that might change your mind.

A Rational Dialogue

I’ve recently been engaged several debates on heated topics in which most participants consider themselves to be rational. Yet, many of these discussions end up with insults being thrown around.

I’d like to offer a few tips, based on my failures and successes, to keep the conversation civil.

Let me first list, from least civil to more civil, statements of disagreement. Let’s use the notion of “rationality/irrationality” as the concept we want to convey.

1. You are a f****** irrational fool!
(Expletives only evoke undesirable emotions.)

2. You are an irrational fool!
(Stay away from nouns of insult.)

3. You are irrational!
(This is stated as a permanent characteristic. Not good.)

4. You are acting/commenting irrationally!
(Better, but why not eliminate the “you”?)

5. Your position is irrational!
(Better yet, but stop yelling.)

6. Your position is irrational.
(Closer to civility, but you’ll need to add reasons.)

7. Your position is irrational as demonstrated by Y and Z.
(Good, but we can make this softer.)

8. It seems irrational to hold X as demonstrated by Y and Z.
(Fine, but “demonstrated” is merely how you perceive it.)

9. It seems irrational to hold X after considering Y and Z.
(Great! But can we now use a more Socratic method?)

10. Isn’t it irrational to hold X after considering Y and Z?
(Great progress, but I think we can do more.)

11. Wouldn’t some consider X to be irrational after considering Y and Z?
(Now we’ve allowed our interlocutors to address the argument without feeling the need to defend their identity. And we’ve innocuously invited them to look at the issue from outside their perspective. Now let’s just add a final few softeners.)

12. Is it possible that some might consider X to be a bit irrational after considering Y and Z?
(Excellent! The apex of civility!)

I’ve written this primarily as a reminder to myself to stay civil and rational.