On Finding Objective Truth

Started by Savant

Replies
24
Posts
25
Page
1 / 1

Conversation

#1 •••

Let's say we want to find an objectively true answer to a question, or at the very least, find a probability distribution for what the likely answer is. There are a few methods we could use to go about achieving this.


1) Intuition


Basically system 1 thinking, which works for a lot of things but isn't always reliable. Probably the most time efficient of all of these methods, but it's also the laziest. Also the most easily distorted by cognitive biases.


2) Opinion Poll


We could take an opinion poll of a lot of people, which has the benefit of a larger sample size but again is often prone to error. For example, if a topic is complex and most people are uninformed, this might just be measuring biases.


3) Poll of Experts


Probably a better approach if we can reliably identify who the experts are, but we could also be measuring selection bias insofar as some people might be more likely to become experts, and then hold opinions they already held. And it's often as difficult to agree on who the experts are and how broad their expertise is as it is to agree on the topic itself.


4) Research


We could research and then use or intuitions based on the new information. But this is the most time-intensive method, and there's the issue of filtering through bad arguments for each side.


5) Debate


We could debate the topic. But this can be pretty time consuming and also depends on each debater being a good representative of the strongest arguments for their viewpoint.


6) Social Platforms


Ideally, this allows for a bunch of people all researching and debating the topic and then figuring out where the majority of people stand. But propaganda, bots, and bias make it far from ideal. As well as most arguments being a few sentences long at most and largely the same ones repeated over and over, which seems inefficient.


So, how can we take the best of these approaches while optimizing them for efficiency?


Phase 1: Argument Mapping


One way to save time might be storing all the arguments in one place so that people don't have to start the argument over from first principles each time. If the chain of argument goes something like point -> counterpoint -> counterpoint, people could be allowed to add arguments at any point along the chain. This could get very big very fast, but if we ask people to rank arguments for the same side against each other, then bias is largely eliminated, since when we ask someone "what is the best counterargument against argument X?" their answer won't really be influenced by whether or not they agree with argument X. So if I want to know the strongest arguments for and against a position, I can just read the strongest argument for, then the strongest counter, then the strongest counter to that, etc.


This system is intentionally adversarial, in that each side is trying to beat the other, yet this does not actually harm the efficacy of the system. If we want to form the strongest argument -> counter chain possible, then it doesn't matter if the CIA is making arguments in favor of Israel and the IRGC is making arguments in favor of Palestine. Insofar as they want to support their chosen side, they will be making the strongest arguments possible for their position, and that's what we want—the strongest arguments for each side. Making bad arguments for the other side would be pointless, since the system is designed such that we read the best arguments, not the bad ones. The only way I could see gaming the system working would be rating bad arguments for the other side high to bury their good arguments, but praising a bad argument for a position you disagree with seems so contrary to human nature that I don't see it being a huge problem.


Phase 2: Acquiring Knowledge from Arguments


So once we have a database of the strongest arguments for each side, what is the best way to derive knowledge from it? Maybe we could have some group read the argument chain and then take an opinion poll, but there's still the question of selection bias. And even if we polled everyone on Earth, is that the optimal sample? If one ideology has more kids than another, might that unfairly skew the poll results?


But there's a better way in my opinion. We sample a bunch of people with preexisting views for either side, ignoring how many of each we get. Then we measure what percent of each group flips their view. Let's say, for example, 2% of people who initially favor the resolution change their mind after reading the arguments, and 1% of people who initially oppose it change their mind.


We then determine what view distribution would be most efficient in terms of information theory—that is, what would the view distribution of the population need to be such that nothing new can be learned from the arguments? In this case, if two thirds of the population opposes the resolution and one third supports it, then the overall poll results won't change after they are exposed to the argument chain—two percent of the smaller group and one percent of the larger group changing their minds will cancel each other out.


To think of this another way, suppose that some small percent of the uninformed population bases their opinion solely on the new information provided by the strongest arguments, and the rest of the uninformed population just keeps the view they had before. Regardless of what that percentage is, presuming it is equivalent regardless of the starting view, we can mathematically estimate the optimal view distribution by sampling each viewpoint group before and after.


Another method might be having people give their percentage certainty in the resolution being true both before and after, then see if they directionally converge toward a specific value. For example, graph people's final certainty vs initial certainty in the resolution being true, and find a line of best fit, then find where the line of best fit intersects y=x. (So for instance, the line of best fit might estimate that if someone is 50% sure the resolution is true beforehand, they are 60% sure afterward. But if they were 80% sure beforehand, then they were 70% sure afterward. But at one point, in this case 65%, we would predict that the certainty remains the same before and after. So we assume an estimate of 65% certainty in the resolution being true is informationally efficient, since on average someone who made that estimate wouldn't change their mind after reading the strongest arguments for each side.)


This phase cannot be adversarial, because it would be easy to skew in favor of one side, but filling out surveys takes less effort and people than making a bunch of arguments, so we could get a sample group of 100 people or so with each viewpoint and verify that they are real people. I think it goes against human nature to voice support for a view one disagrees with, so I think most people will be honest.


Basis for Efficiency


The theory behind this system isn't that it always arrives at an objectively true answer, but that it arrives as close to the objectively correct answer as the population possibly can—that is, there are no additional sources of evidence, arguments, or intuitions that can be exhausted once this system has been used that would aid in reaching on objectively true answer. In practice, people won't submit every possible argument and response, but they might get close enough to be effectively efficient.

Edit comment

#2 •••
@Savant

How would you deal with views changing over long periods? Ancient cultures were ok with man boy love, how does or would that be evaluated? This point is often brought up when arguing for non objective truth so how does it factor into your system


Is your system just majority rule? If not how do you credit arguments in the minority as still true? Just reading the options and risking a position?

Edit post

#3 •••
@linate

It's not majority rule, it's evaluating which position people move towards when presented with the strongest arguments for each side. The system doesn't promise to be correct but only efficient; that is, it exhausts all available information. Much like the price of a stock reflects all available information even though it changes over time.

Edit post

#4 •••
@Savant

Do you think the long term goal thousands of years would likely lead to more objective truth much like how stock prices generally go up... or would the truth of propositions merely fluctuate all over the place as cultures evolve and change ?

Edit post

#5 •••
@Savant

If propositions merely move all over the place, it wouldnt so much be reaching towards truth, so much as merely observing and recording beliefs as they evolve


It would be good for forming informed beliefs but maybe not so much objective beliefs. Is being informed the closest we can get to objectivity even if the objectivity cant be established? Isnt the whole point of this exercise to establish objectivity ?

Edit post

#6 •••
@Savant

Finding objective truth simply means starting from logical law as a premise, and as long as someone accepts that logic even exists and isnt too stupid, they must accept premise 1 because premise 1 is merely a basic logical law.


Good example of premise 1 is:


P1. A is true or Not A is true


there is no way to disagree with this premise, without denying that logic even exists.


So as long as someone accepts that logical laws exist, any first premise made in form of logical law itself must be accepted as true.


Another example of objective truth is definitional conditional premise:


P1. If definition of A is B, then A is B


this premise remains true by law of identity.


third, weaker form of objective truth is observational premise.


P1. We see that A happens with B


But this one can be denied by questioning validity of observations



But objective truth isnt gained by debating, or polls, or "experts". these entirely depend on who decides what truth is. A perfectly reasonable argument can be rejected, while unreasonable one may be accepted, depending on who the "judge of truth is".


the main problem in determining truth is that such right cannot be exclusive to anyone. In same way reciprocity works with morality, it likewise works with truth determining. If you say "We decide what truth is", I can simply say "No, I do.".

Edit post

#7 •••
@linate
If propositions merely move all over the place, it wouldnt so much be reaching towards truth, so much as merely observing and recording beliefs as they evolve

I'm not sure it would fluctuate. So long as the strongest arguments don't change, the initial beliefs of the population being measured don't matter.

Edit post

#8 •••
@Savant

Also, here you are starting from the premise that people actually care about whole truth. Now that premise is difficult to prove.

Edit post

#9 •••
@SatanLucy
Also, here you are starting from the premise that people actually care about whole truth. Now that premise is difficult to prove.

People are more likely to shift toward the truth than away from it in the presence of new information. Otherwise how do you ever learn anything? You'd simply have to assume that any time you learned something, it's just as likely you became more incorrect as that you became more correct and thus you have zero information.

Edit post

#10 •••
@Savant
People are more likely to shift toward the truth than away from it in the presence of new information.


No. the presence of mass amount of information merely drowns the truth. Its like overloading a hard drive with bunch of data. It just stops working. taoism is fully correct when saying that talking too much merely muddles the water. If truth exists as bunch of data which cannot ever all be had or even remembered, full truth can never be reached at all.


Otherwise how do you ever learn anything?


the masses dont really learn. If they did, you wouldnt need to make this topic in the first place. they would already have objective truth, and topic would be meaningless. they dont.


the whole problem of your idea is amazing ability of masses to forget. If you give masses bunch of information, they will remember almost none of it. Its why 12 years of education produces idiots. Because human brain, being limited in capacity, cannot learn anything without forgetting something else when capacity is full. Its simply not logically possible to add more data to full hard drive without deleting something. And if you add too much, the limited hard drive merely rejects what cant fit.


Second problem is that you start from assumption of honesty, in a world which is dishonest to the core. the people dont even want the truth. they want their interests, and those change a lot. If people cared about truth, liars wouldnt exist, and people would question more instead of mindlessly swallowing same propaganda type each time.


So reaching some objective ultimate truth is not possible for most people, not even for most experts. It is only few of the few who have the truth, and the rest are NPC (non-player characters) who just accept whatever "feels good" at a time.


Most people dont even think in terms of formal logic, so claiming they can reach logical truth while not knowing logical laws is like saying you can win at chess without knowing where pieces can be moved and where not.

Edit post

#11 •••
@SatanLucy
the masses dont really learn

Really? Never? So people don't speak any languages, or build things? Or do anything they didn't already know how to do as a two year old?

Edit post

#12 •••
@Savant
Really? Never? So people don't speak any languages, or build things? Or do anything they didn't already know how to do as a two year old?


For each new thing, some previous one is forgotten when brain is at full capacity. This is simply logically true.


So for each new thing you teach the masses, they merely forget something else. For example, people in Japan forgot how to reproduce. It happens.


What is also logically true is that even if person was reading 24 hours a day each day for the rest of his life, he wouldn't be able to read even 10% of knowledge out there in the world.


So it is logically impossible for average person to even reach 10% of truth.


For AI it may be possible. For human, nah.


It's why Tao teaches that the only realistic truth is core truth, and countless data is not a good way to truth.

Edit post

#13 •••
@SatanLucy

So they will learn facts relevant to the specific issue being debated. Still good enough.

Edit post

#14 •••
@Savant
So they will learn facts relevant to the specific issue being debated. Still good enough


Maybe it's "good enough". It's not going to guide them to any objective all-truth. There is a reason why some people are called "experts" in some fields. Because other people are not. And even experts miscalculate, so what hope is there for average person who is not expert on anything?

Edit post

#15 •••
@SatanLucy
It's not going to guide them to any objective all-truth.

It doesn't have to, the method just assumes it gets them closer to the truth. And determines what value everyone is getting closer to.

Edit post

#16 •••
@Savant

"On Finding Objective Truth"


If it's merely about getting closer to it, sure there are ways. But getting closer to the river while never getting to the river isn't really much to hope for.

Edit post

#17 •••
@SatanLucy

Well, the process of finding objective truth involves getting closer to the truth. In this case, if we presume that 100 different people are getting closer to the truth, then we can become a lot more accurate than all of them by determining what value they are converging toward. If everyone's estimated probability of a proposition being true is getting closer to 70%, then 70% is our best estimate.

Edit post

#18 •••
@Savant

How about this, ask an average person to list you basic laws of logic. If they can't do that, then any thinking they do is essentially feelings based. And this is before you even ask them how logical reasoning works.


They simply can't reach objective truth if they are not even using logical tools to reach it.


Logical reasoning has rules, and how much average person even knows about them?

Edit post

#19 •••
@SatanLucy
They simply can't reach objective truth if they are not even using logical tools to reach it.

That's the point of the argument map...to give them the logical tools.

Edit post

#20 •••
@Savant
That's the point of the argument map...to give them the logical tools


Argument map is for people who already know logical laws. For people who don't know logical laws, judging arguments is impossible and is entirely feelings based.

Edit post

#21 •••
@SatanLucy
For people who don't know logical laws, judging arguments is impossible and is entirely feelings based.

They're not judging the arguments, simply giving their beliefs before and after. Intuition and feelings are kind of the point anyway, it's the only thing the arguments themselves don't cover and you have to start from some amount of intuitions to use as premises, otherwise the arguments are meaningless.

Edit post

#22 •••
@Savant
They're not judging the arguments, simply giving their beliefs before and after. Intuition and feelings are kind of the point anyway


So now it's "feelings"?


Well, I don't see how objective truth is reliably produced by someone's feelings.


As far as I know, when feelings start, that's usually where thinking ceases.

Edit post

#23 •••
@SatanLucy
I don't see how objective truth is reliably produced by someone's feelings.
As far as I know, when feelings start, that's usually where thinking ceases.

How do you justify "when feelings start, that's usually where thinking ceases"? You can use some logic to get there, but you'll need starting premises, and you have to use a priori beliefs to get those.


Also if the feelings are just bias, they will be baked into the "before" belief...the whole point is to see how their belief shifts to before vs. after reading the argument.

Edit post

#24 •••
@Savant
You can use some logic to get there, but you'll need starting premises, and you have to use a priori beliefs to get those.


If someone believes that logic exists, they are already accepting all formal logic laws as possible starting premises, assuming they know them.


But if someone doesn't know what logic even is, how can that person possibly use logic to judge arguments?


Because you don't solve math equation by feelings.


And solving countless arguments in order to decide about belief change?


Belief change based on feelings can also be due to rhetoric. Because "feelings" aren't exactly famous for being logical.

Edit post

#25 •••
@SatanLucy

Logic needs a sound and reliable knowledge base to test if our ideas are correct.

Edit post