Of course, it is obvious! Subjectivity, (in contrast to objectivity), drive human behaviour.

I thought objective thinking is a virtue, because we thought bias is wrong, because, by being objective; justice will be served, important quality for being a judge.

But I think that will hinder problem solving skills.

Do you know Occam's Razor? it state "in explaining a things, no more assumptions should be made than are necessary."

If we tried to follow Occam's Razor, and at the same time being objective, and then asked to explain things using a given limited amount of information or clue (like patient's symptom in Dr. House); then we are bound to made error.

The error are made because of hasty conclusion.

EDIT: It was hasty because there's no further motivation to gather information or clue. No pre-assumption was made to be proven.


Since we are human, we are subjective. Instead of evaluating, we made hypothesis and justify them with observed symptom. It is like "we understand what we want to understand".

But it is better than that!; we appear to be able to be objective by accepting negative or null confirmation and move on to next hypothesis.

but still... that's either because we were being objective (for a moment), or because we always had hypothesis for anything which leaped for one to another, like frog (subjective forever). For example; we use divine illness to explain the symptom which cannot be explained by other mortal desease after exhaustive search.



!amazed Of course.... !amazed if we were to send an AI to other planet, it wasn't just mere intelligence we should accounted for, it should be this subjective capability!

Maybe there should be several computer working independently, each with unfair bias to certain concept, yet still cooperative and not act irrationally.

But, x) of course... x) human weighted some concept differently because of experience... that mean putting numerous computer with different bias in a trial and see which yield sensible result.

and... %&*....Ok... the sensible result is to send man. All social function and emotion that proven human functionality in group are there already, no tweak is needed.

EDIT: maybe even humans need to be finely tuned and paired.



I watched "From Earth to the Moon", in Apollo 17 a geologist were sent to the moon. A geologist/astronout observed a sun reflection on the soil, which is from the gold foil on the lunar rover and immediately excited on the COM with the finding of yellow soil (which prove something). However, that's clearly wasn't a yellow soil, he realize that. But eventually, he found a yellow soil at other time and place on the moon. The only Apollo mission with a scientist onboard. I think that was a good example.

EDIT: If there is no geologist in Apollo 17, maybe they didn't even look for yellow soil.



Reference: Dr. House episode 11 season 1 and From Earth to the Moon part 10: Galileo Was Right.

lol

en.wikipedia.org/wiki/Construal

Lee D. Ross is a professor of social psychology at Stanford University is currently working on his theory of a special kind of construal which he calls "naïve realism." In a simple experiment, Ross takes peace proposals created by Israeli negotiators, labeled them as Palestinian proposals, and told Israeli citizens that the ideas on the proposal were the ideas that Palestinian wanted the Israeli to adopt. Then took the same proposals and left them the same way and told the Israelis that ideas on the proposal were the ideas that the Israelis wanted the Palestinians to adopt. The Israeli citizens liked the proposals from the Israelis to the Palestinians more than the proposal from the Palestinians to the Israelis even though they were the same proposal.

"Even when each side recognizes that the other side perceives the issues differently, each thinks that the other side is biased while they themselves are objective and that their own perceptions of reality should provide the basis for settlement."

I had an insight recently, of semi-relevance.
All negative judgements are hypocritical in essence.
We judge others for their bad attitudes, and while doing so have a bad attitude about it. We are negative about other peoples negativity. "That jerk is so negative!"

Because we are biased to see negative things?

It seems that bias is anything about human behaviour.

I think it should be interesting to have a bias AI.

[quote1231274509=MAGNON]
I think it should be interesting to have a bias AI.
[/quote1231274509]
Hmmmmmmmm. Interesting thought.
AI has to go according to its programming. Its programming is a bias, in a sense.
On the other hand, if bias is an emotional rather than mental thing, then an AI could not be biased.

Can an AI be given objective standards of logical judgement?

Errr.... that make sense, human intelligence and animal intelligence is observed based on their peculiar behaviour.

But, the main theme of this thread is that; bias is causing human behaviour.

In wikipedia, the AI is defined, by the prominent figure in AI research, as the 'selective input/output of the brain'. So, that's the bias.

By integrating those quality into a logic software, we are indirectly integrating human behaviour into those logic software. Perhaps abit, or perhaps much, it depend on how complex the selection criteria.

IF so, then it wasn't interesting at all to have a bias AI, since an AI is the definition of a biased logic software. The best AI would then be human.


I think an AI cannot have those "objective standards of logical argument". It can, only if, the software evaluate input of 1 or 0 (like switch, or in human comparision; like a single nerves). However, the brain already 'conditioned' 1010101 into different value... and an AI would do the same too.

To 'condition' the mixed 1010101 would need an AI, a logic software that look into the input for a recognizable pattern and ignore other as noise. That's bias, right?

So, no AI can be objective.


Hmmm... I don't know what is "The Standard of Logical Argument". (Perhaps it was a philosophical term?) But, IF that mean we want an AI to comply to human 'standard', then it make sense.

Thanks.