[Automatically Generated]

00:00:38 Speaker 1
Welcome to reasonable Frippery, a podcast from the Carl-Friedrich von Weizsäcker Center in Tübingen, where we explore the known and the unknown with scientists and experts from various fields.
00:00:48 Speaker 1
I'm Philipp, your host, and this is the very first episode.
00:00:52 Speaker 1
I'm joined by Vlasta Sikimic, an expert in the field of epistemology.
00:00:57 Speaker 1
And together we will talk about what is knowledge to social view and knowledge acquisition, the difference between knowing that and knowing why the role of justification in gaining knowledge prediction versus explanation, the borders of the knowledgeable scientific models as approximations of truth and differences in individual versus social knowledge.
00:01:18 Speaker 1
Position Lester will be introducing herself and providing her unique perspective on these important topics.
00:01:24 Speaker 1
So sit back, relax, and join us for thought provoking conversation on reasonable frippery.
00:01:31 Speaker 1
Today we are going to talk about topics that revolve around epistemology, and we will also specifically talk about your research and saviers.
00:01:41 Speaker 1
But before we start with the actual contents, it would be great if you could quickly introduce yourself and maybe mention the things that made you who you are in a sense, right.
00:01:50 Speaker 1
The things that brought you into the field of epistemology so stage is yours.
00:01:55 Speaker 2
Thank you so much Philip.
00:01:57 Speaker 2
I am very happy that we are making this podcast today.
00:02:00 Speaker 2
I am.
00:02:01 Speaker 2
My name is Lacey Kimmich.
00:02:03 Speaker 2
I am a researcher in the domain of philosophy and I mainly work on social epistemology of science with a huge interest in social epistemology in general.
00:02:15 Speaker 2
And I also teach this subject at the University of tubing and this semester.
00:02:21 Speaker 2
My background is I'm very strong in the formal competency, so I have a background.
00:02:29 Speaker 2
Logic and also worked on data-driven mechanisms in science, which means applications of machine learning algorithms in science. So what brought me to this point in life is that I was of course.
00:02:49 Speaker 2
Very interested in many varied topics during my school years, I was going to this mad competitions which is very popular for the part of the world where I come from and I come.
00:03:03 Speaker 2
From this we can call it Eastern Pike.
00:03:05 Speaker 2
Even though you can Slavia means Southern Slavic country.
00:03:09 Speaker 2
So I come from Serbia, but I was actually born in this country that doesn't exist anymore and that was called in Yugoslavia.
00:03:18 Speaker 2
And so yes.
00:03:20 Speaker 2
So so of course there are these very positive aspects kind of how one.
00:03:24 Speaker 2
Went to these competitions, got awards in mathematics, study logics in Amsterdam was a really good during the studies and so on.
00:03:35 Speaker 2
However, I always like to point to my.
00:03:37 Speaker 2
Students that none of these things are easy, and I always try to encourage them and to tell them that also their lecture nowadays had a lot of problems before and to to give you some illustrations.
00:03:50 Speaker 2
So there are quite some students that I now have that are very nervous before presenting because of their English skills.
00:03:58 Speaker 2
And I always tell them they'll also the first presentation I gave in English that happened during my Masters studies, it went fine, but it didn't excel by no means.
00:04:09 Speaker 2
It's more like it passed and.
00:04:13 Speaker 2
And it's also, it's also important to get to kind of know that all of it is also a lot of hard work and not all the doors open to one person.
00:04:22 Speaker 2
And being a woman interested in these subjects was not always easy, and they're also use one example that when I was the one lovely conference in logic in Tubingen.
00:04:34 Speaker 2
And it was really a great experience.
00:04:36 Speaker 2
Everyone was extremely friendly and that was quite some years ago.
00:04:40 Speaker 2
But I was the youngest person in the room.
00:04:43 Speaker 2
I was only attending, not presenting, and I was also the only female person in the.
00:04:49 Speaker 2
And of course, I didn't feel so nice, even though everybody was super friendly.
00:04:54 Speaker 2
And I think it's important kind of this, this message that we all fought a lot through our lifetime to, to, to reach the position where we are.
00:05:04 Speaker 2
And a lot of doors don't open immediately like we have to knock many, many times.
00:05:10 Speaker 2
And only then they will open.
00:05:12 Speaker 1
Thanks a lot for this quick intro.
00:05:15 Speaker 1
And and yeah, as I said in the beginning today we are going to talk about epistemology.
00:05:21 Speaker 1
And before we start with answering or exploring any specific questions, it would be fantastic if we just get clear of what we mean with epistemology.
00:05:32 Speaker 1
So maybe you can start with the.
00:05:34 Speaker 1
All notion what is epistemology about?
00:05:37 Speaker 2
So epistemology is we can also call it theory of knowledge.
00:05:42 Speaker 2
So the way how we learn, how we gain knowledge or in other way way.
00:05:46 Speaker 2
In other words, how we learn and this social take on epistemology is how we learn in social settings or how we learn from.
00:05:54 Speaker 2
Other people, we can also think how we learn from AI and from from different engines.
00:06:02 Speaker 2
Maybe one nice popular example of kind of to understand what would the social epistemology be, or in general, what the epistemology deals about would be this recent movie don't look up, so it's kind of how people communicate expert knowledge, how they accept the knowledge so they have enough patience to listen.
00:06:22 Speaker 2
To experts, how is elitistic approach to science affecting public knows?
00:06:28 Speaker 2
Much and this knowledge of course, makes huge impact on our lives and our decisions.
00:06:35 Speaker 2
And that's why it's so important and fascinating.
00:06:38 Speaker 2
But the traditional some of the more traditional topics in epistemology, we're dealing with these scenarios of whether we are in some sort of metrics, whether we are systematically deceived, whether reality is is it appears to us.
00:06:51 Speaker 2
So there are many different topics that epistemology.
00:06:56 Speaker 2
Deals with at the moment.
00:06:59 Speaker 1
So epistemology is not about or.
00:07:02 Speaker 1
It's maybe also about distinguishing the facts that we can learn about the world that are true from the facts that are.
00:07:11 Speaker 1
Non true, I think facts are always true, right?
00:07:13 Speaker 1
I think that's a characteristics of facts.
00:07:16 Speaker 1
But I mean this notion of things that we learn.
00:07:19 Speaker 1
We we try to.
00:07:20 Speaker 1
To differentiate properly in the space of epistemology from the right and from the wrongs.
00:07:26 Speaker 2
Yes, definitely.
00:07:28 Speaker 2
So there is this traditional debate kind of what is knowledge and how can we form real knowledge instead of.
00:07:37 Speaker 2
Having false beliefs.
00:07:39 Speaker 2
And there it's usually said through proper justice.
00:07:44 Speaker 2
But we also.
00:07:44 Speaker 2
Have different approaches like virtual epistemology which tells us which predispositions we need to have in order to learn most or to gather knowledge in the best way or how.
00:07:56 Speaker 2
How should we be like for instance, we should be epistemically open to different arguments, and that is a very good.
00:08:03 Speaker 2
Predisposition of learning something that's.
00:08:09 Speaker 2
More truthful, or that's truthful.
00:08:11 Speaker 2
That's always the problem.
00:08:12 Speaker 2
How to say it?
00:08:13 Speaker 1
Yeah, it's quite interesting.
00:08:14 Speaker 1
So epistemology by itself also covers many different scientific disciplines, right?
00:08:20 Speaker 1
Because learning is about the cognitive sciences.
00:08:25 Speaker 1
And I think knowledge in itself could be, I don't know, a purely philosophical topic and knowledge includes also more than only the cognitive sciences, so it it.
00:08:35 Speaker 1
Seems to be.
00:08:36 Speaker 1
Very broad and it's capturing a lot of different perspectives on the quest of gaining knowledge, I guess.
00:08:43 Speaker 1
That's that's quite interesting.
00:08:45 Speaker 1
So in epistemology there is the question of or there is the phrase of justified true belief, and I was wondering what is this notion about and how does it relate to learning?
00:08:56 Speaker 2
So that's Plato's definition of knowledge, the most famous one which says that in order to to have knowledge, we have to have a justified true.
00:09:08 Speaker 2
Leave, and this was later disputed by Gautier and Gautier. It's a very, very famous debate in epistemology about Gauthier's counter examples which tell you you can have a justified true belief, but you still don't have knowledge. One of the paradigmatic example is a person.
00:09:29 Speaker 2
Folks said.
00:09:31 Speaker 2
Watch and the watch stopped, but by coincidence the watch is showing 12:00 o'clock, which is really the the case.
00:09:40 Speaker 2
So it's showing the accurate time but by coincidence and then the person sees it and forms a belief that it's 12:00 o'clock this belief.
00:09:50 Speaker 2
Happens to be true and even has a justification.
00:09:53 Speaker 2
I just looked at.
00:09:53 Speaker 2
The watch, so that must be the case, but we would still tend to say that this person doesn't have knowledge.
00:10:00 Speaker 2
Because this is a consequence of a mere coincidence, and in this way then a lot of.
00:10:07 Speaker 2
Other theories were developed in order to try to explain how can we really have knowledge and how should we define knowledge.
00:10:15 Speaker 2
And those are these more traditional questions in epistemology which.
00:10:20 Speaker 2
Are now changed or not really changed but they are getting slowly shifted by these other questions of how do we have?
00:10:28 Speaker 2
How do we learn?
00:10:29 Speaker 2
So how do we learn within a group?
00:10:31 Speaker 2
How do we get this misinformation spread?
00:10:34 Speaker 2
How do we get information?
00:10:36 Speaker 2
And for my taste, these more contemporary questions are more interesting.
00:10:43 Speaker 2
But there are also some of these traditional questions which are really instructional and insightful.
00:10:49 Speaker 2
For instance, between the distinguishing between knowing how and knowing that, that's that's, I think, something.
00:10:58 Speaker 2
Knowing how and knowing that would, for instance, be very good indications of a person who really knows a lot of grammar of the foreign language and of a person who is actually able to speak the foreign language.
00:11:12 Speaker 2
So at my school I really learned a lot of grammar with respect to in the German language.
00:11:18 Speaker 2
I could solve a lot a lot of tests with the grammar in German, which is not trivial.
00:11:25 Speaker 2
However, I'm not.
00:11:27 Speaker 2
Well, I'm getting better and better in speaking, but definitely ever.
00:11:30 Speaker 2
The child who speaks fluently German and doesn't know these grammatical rules is far more competent in the language, and the child knows how, and I maybe know in some context that something should be the case.
00:11:46 Speaker 2
Similar things would be said, kind of someone can know theoretically how body.
00:11:51 Speaker 2
Moves when we were swimming but was never in water, swimming while other people just swim and don't know the physiological processes behind it or these explanations with with muscles and I I also don't know them so.
00:12:06 Speaker 2
That, that those those two distinctions can be insightful when we think about certain how we explain certain, whether we know how to explain something, whether we know how something functions.
00:12:20 Speaker 2
So whether we have an explanation or?
00:12:23 Speaker 2
We only can.
00:12:23 Speaker 2
Use something.
00:12:24 Speaker 2
So maybe we have some useful prediction or.
00:12:27 Speaker 2
Knowledge is very usable and applied, or whether we can really explain some mechanisms behind certain behavior.
00:12:35 Speaker 1
You have mentioned on the other hand, the justified notion and you said for example in the watch watch example that you gave right where you sometimes look at the watch and by coincidence the other watch also shows 12:00 o'clock that you form a justified true belief which is true, yes.
00:12:55 Speaker 1
But the way it was justified was rather superficial.
00:13:00 Speaker 1
So although the person who forms the belief may consider the thing that happens.
00:13:07 Speaker 1
It could also be that it's not proper knowledge since the justification was so superficial and I was wondering whether this notion of justification is.
00:13:19 Speaker 1
Rather, to be seen in terms of degrees than in terms of categories, right.
00:13:25 Speaker 1
So when we form a category that somehow says yes, this is just.
00:13:29 Speaker 1
Right or no, it is not justified. We can also think about forming like levels of degrees, right? So looking at the watch one time and seeing the coincidence at the other watch is maybe not so much justified as by looking at the watch 1000 times.
00:13:49 Speaker 1
You know, and doing this comparison 1000 times and then form a justified true belief. So what is?
00:13:57 Speaker 1
Justification about and is it really something that can be seen in terms of categories or is it something that should be seen in terms of degrees?
00:14:06 Speaker 2
That's a very, very interesting point and thank you for it.
00:14:11 Speaker 2
And with respect to these traditional answers, we could say, for instance, this the justification didn't start from the right premises or kind of the derivation of the of the proof was was not the one that corresponds to reality.
00:14:25 Speaker 2
It's on.
00:14:26 Speaker 2
However, what you are aiming for is something that scientists do every day, so they have to kind of experiment in proper ways because the the way how we are learning about the reality and in science is of course in a way imperfect and.
00:14:46 Speaker 2
There are these.
00:14:47 Speaker 2
Strict scientific methods that should help us to come closer to the truth, and that's one of the examples.
00:14:54 Speaker 2
Paradigmatic examples is exactly what you just mentioned, is kind of to repeat the experiment in different.
00:15:02 Speaker 2
So in this case of the watch, look at it at different time and see what happens.
00:15:07 Speaker 2
And that's in a way, this replication of the experiment.
00:15:11 Speaker 2
So that's that's something that when we think of the justification in the scientific context in the context of experimenting.
00:15:21 Speaker 2
Of course that we can have more reliable results and less reliable results, and the more scientists replicate some experiment, the more robust the results are.
00:15:34 Speaker 2
So that's a that's an excellent point.
00:15:37 Speaker 1
Another I'm questioning disregard justification seems not to be an end in itself.
00:15:42 Speaker 1
We want to achieve something with the things that we have justified, right.
00:15:46 Speaker 1
Knowledge is not an end in itself.
00:15:48 Speaker 1
We often gain knowledge in order to do better predictions in our environment, or to do better decisions.
00:15:56 Speaker 1
And I was wondering.
00:15:58 Speaker 1
And what is the interplay of these components?
00:16:00 Speaker 1
Right, and why knowledge is so important for decisions?
00:16:03 Speaker 1
And isn't it a like inherent quality of knowledge to say something about our world to make proper predictions about what is going to happen next?
00:16:14 Speaker 2
Yes, I would fully agree with this.
00:16:17 Speaker 2
That knowledge serves for us to to make decisions to make informed decisions and that we can still not listen to knowledge but listen to emotions and to different aspects when making decisions, however.
00:16:34 Speaker 2
This idea of predictions is something very interesting for philosophy of science.
00:16:41 Speaker 2
And it's interesting exactly in this respect, when we are considering whether we are satisfied with a prediction or we want an explanation and all these approaches with machine learning could give us predictions.
00:16:57 Speaker 2
But now there are ideas that machine learning could maybe even provide certain types of explanations.
00:17:02 Speaker 2
But that's traditionally considered to be difficult.
00:17:06 Speaker 2
And and then kind of what is a prediction without an explanation?
00:17:10 Speaker 2
And and all sorts of these questions.
00:17:12 Speaker 2
So are we interested in how is something possible to happen or are we only interested in what is going to happen or what we could do with these counterfactual models, kind of what we could do to change?
00:17:26 Speaker 2
Something in the future, so that's that's all very, I mean those topics are very timely and there is also rather rich literature on them.
00:17:37 Speaker 2
And I think it will the field will continue to expand in this direct.
00:17:41 Speaker 2
And when it comes to decisions, there is, of course always this gap kind of how do we come from knowledge to action and how does someone because the action also requires certain motivation and so on.
00:17:55 Speaker 2
And in this idealized rational agents, they will just follow what they know.
00:18:01 Speaker 2
They will just do the best.
00:18:03 Speaker 2
Rational thing, but it's not what humans always do.
00:18:07 Speaker 2
However, gathering knowledge should of course help us to make, as I said, this informed decisions.
00:18:14 Speaker 2
And interesting aspects are also these theories of epistemic democracy.
00:18:21 Speaker 2
Which kind of tell us when we get as much as possible knowledge, then we will make decisions that will be very kind of in a democratic society that will lead us in the best possible way that will lead us to truth.
00:18:35 Speaker 2
And that's of course a huge question whether something that we vote.
00:18:41 Speaker 2
Or should discover us truth, or that should be kind of containing our values and what's the interplay between?
00:18:49 Speaker 2
And knowledge and.
00:18:51 Speaker 1
That's indeed quite interesting.
00:18:54 Speaker 1
All right, let's zoom out a little bit.
00:18:57 Speaker 1
Is there knowledge that we can?
00:19:00 Speaker 1
Not identify in principle because I was wondering I'm I I know that you're also interested in AI or philosophy.
00:19:09 Speaker 1
And uhm.
00:19:12 Speaker 1
I somehow get the feeling that this black box issue that we have in AI is maybe an issue that cannot be solved in principle, so I was wondering could be complexity or could be connectivist stics systems like the brains or like.
00:19:33 Speaker 1
Deep learning Nets.
00:19:34 Speaker 1
Could this be a thing?
00:19:35 Speaker 1
That is in principle.
00:19:38 Speaker 1
Not fully explainable like is there maybe already a border that we cannot pass when it comes to gaining knowledge?
00:19:48 Speaker 1
Do you have intuitions about this question?
00:19:50 Speaker 2
So I'm not sure so maybe we can think about these laws of physics that kind of these laws on indetermination that we already don't know something and we can never know, you know, what happens there.
00:20:01 Speaker 2
Then there are these ideas that I recently read about this futuristic idea, which we don't know whether it.
00:20:07 Speaker 2
Will be true but.
00:20:09 Speaker 2
That laws of physics might be changed.
00:20:11 Speaker 2
So there are a lot of strange ideas, but philosophers they would already tell something different.
00:20:17 Speaker 2
They would say this skeptical argument that we cannot, that we only know for certain this consciousness and that we know this laws of logic and mathematics for certain, but we we cannot ever eliminate.
00:20:30 Speaker 2
This skeptical scenario, so they would even argue so these radical skeptics, they would even argue that we don't know with certainty anything about the outside world.
00:20:43 Speaker 2
So that we only know it with some degree and that we can never prove the existence of the outside world, which is even more radical.
00:20:52 Speaker 2
But that's kind of, uh, what they would say even nowadays.
00:20:56 Speaker 2
And you know this Humes gap and so on.
00:20:59 Speaker 2
But I I mean it's it's it's a really fascinating question.
00:21:02 Speaker 2
I just I can only so the problem is I can only speculate about it.
00:21:06 Speaker 2
I cannot really properly answer it an expert.
00:21:11 Speaker 2
But I think it's a great topic.
00:21:12 Speaker 2
I think if if someone would write a good article either for the general public or even better scientific that that would like a lot of people, would react to it because the topic.
00:21:22 Speaker 2
Is good and it's hard.
00:21:25 Speaker 1
Yeah, I I think it it is definitely.
00:21:27 Speaker 1
Hard and I I.
00:21:28 Speaker 1
Somehow buy in or I buy this perspective of that all that we discover all the scientific models that we have are.
00:21:36 Speaker 1
Can only be considered as mirror approximations of whatever of, of whatever the thing may be that we call truth.
00:21:45 Speaker 1
And this quality of being an approximation of something means also that there is potential to find better approximations of this thing that we call truth.
00:21:55 Speaker 1
But I was wondering.
00:21:59 Speaker 1
Could there be a potential point at which we could say that our approximation is now so good that we can call this approximation the absolute truth and I know absolute truth?
00:22:11 Speaker 1
Let's not linger on this word a little bit too much, but I was like in a thought experiment thinking about could there be a state?
00:22:19 Speaker 1
Of knowledge of which we can say, hey this.
00:22:23 Speaker 1
Is really.
00:22:27 Speaker 1
Very, very, very, very good approximation.
00:22:30 Speaker 1
So following thought experiment when we are capable to come up with physical descriptions of the world that answers all our chains of wishes that could be potentially answered so.
00:22:46 Speaker 1
So it's like the Laplace demon argument.
00:22:48 Speaker 1
I don't know whether you've heard about this.
00:22:50 Speaker 1
If you know the position and the movement of each particle in the universe, then you can predict all states of the universe.
00:22:56 Speaker 1
If it is deterministic from the beginning until the and the idea is if you could create a model that can do this stuff and we have the computational power to simulate all of this.
00:23:09 Speaker 1
And we have no gaps in our chains of Y.
00:23:11 Speaker 1
Any more then I would consider this model that we have of the world as a super good approximation if not.
00:23:21 Speaker 1
The truth then, in the end, you know if if every prediction that we do about the universe is 100% true.
00:23:29 Speaker 1
Like or is actually like being predicted as a future state of the universe.
00:23:35 Speaker 1
Then the sort of knowledge that we have gained could be actually the really true knowledge and and that that was, you know and and this is then.
00:23:44 Speaker 1
For me, a state where.
00:23:49 Speaker 1
We could have.
00:23:51 Speaker 1
You know all.
00:23:52 Speaker 1
Truths gained in this world, or sorry, all knowledge potentially gained in this world being actually gained, being well aware of that.
00:24:00 Speaker 1
There are wise that we maybe in principle cannot answer.
00:24:05 Speaker 1
For example, what happens before The Big Bang, right?
00:24:09 Speaker 1
So maybe time there is no before there.
00:24:13 Speaker 1
It's like an unreasonable question.
00:24:15 Speaker 1
Like what is?
00:24:19 Speaker 1
The temperature lower than the absolute zero temperature right there is nothing there.
00:24:24 Speaker 1
It's a physical nonsense question, but you can ask this, but there might be no answer to it.
00:24:29 Speaker 1
So these sorts of.
00:24:32 Speaker 1
Questions may exist also in this model where we can explain almost everything.
00:24:36 Speaker 1
You know that we know, OK, there is there are certain things which cannot be answered in principle, but also this I consider as knowledge, right knowing that questions cannot be answered in principle, yes.
00:24:46 Speaker 2
Sure, limitations.
00:24:47 Speaker 2
Yes, but so and how do you defend against gettier's counterexample? That's kind of all. This is just by coincidence. True.
00:24:59 Speaker 1
I think I don't know whether it's hold. It's holding still in this thought experiment because if you think about having a model that really predicts 100% what happens in the future, right?
00:25:10 Speaker 1
You, as a human can look at your computer and you can say you're nodding in the next second or I don't know.
00:25:19 Speaker 1
Zabina in Munchen or Munich throws a glass.
00:25:22 Speaker 1
Now you know you can predict every state of the universe.
00:25:26 Speaker 1
And if you have this sort of model, you know, then it's not a coincidence anymore, but in. On the contrary, it's a 100% certain prediction.
00:25:37 Speaker 2
And how do you defend against this evil demon?
00:25:41 Speaker 2
So this radical skepticism that you're just in the matrix that you are just?
00:25:45 Speaker 2
Fed this information.
00:25:48 Speaker 1
Yeah, and this is again about the wines that we cannot know in principle.
00:25:52 Speaker 2
OK.
00:25:52 Speaker 1
So if we are in a virtual environment, right?
00:25:57 Speaker 1
And when we now build approximations toward this virtual environment, it could still be that there is something outer.
00:26:07 Speaker 1
That is beyond this virtual environment, right?
00:26:09 Speaker 1
That is simulated somehow, and these truths I would consider as being not answered in principle.
00:26:17 Speaker 1
Because they are beyond the thing that we are currently.
00:26:19 Speaker 2
Approximating Sir and how do you guarantee that these predictions will continue to hold over the time?
00:26:27 Speaker 2
You know, this assumes argument kind of the the fact that these predictions were accurate 100% until now all the time. How do we?
00:26:35 Speaker 2
Know that you know the sun will rise next morning.
00:26:39 Speaker 1
Yeah, uhm I think I mean it.
00:26:43 Speaker 1
It may not be possible.
00:26:45 Speaker 1
I except I I mean if we.
00:26:50 Speaker 1
Could do it. I mean, if we have a computer program who is doing it for the last 100 years with the Hun 100% correct manner, and if we and even can look back in the past, that's what also Laplace demon predict if we live in a deterministic universe, we cannot look just in the past, but also in the future.
00:27:10 Speaker 1
It would be also a scientific thing.
00:27:13 Speaker 1
I don't know whether it's truly scientific.
00:27:14 Speaker 1
But if our.
00:27:15 Speaker 1
Observation is always 100% mapped to the model that we have developed, then this might be the empirically perfect model, right?
00:27:26 Speaker 1
Because it predicts all the states of the past to 100% and all the states of the future, potentially in 100%, although we not.
00:27:33 Speaker 1
So it still would be, I would think the the most scientific and best approximation of what happens there.
00:27:42 Speaker 1
So I think it's not defendable.
00:27:45 Speaker 1
I mean, we do not know yet what's coming in the future, but again, with this model, assuming you know it could be that we have maybe at some point so much knowledge gained that we can maybe predict to.
00:27:57 Speaker 1
To 100% what's happened in the future?
00:28:00 Speaker 1
There are no things that intervene with it or like, for example, it could be that in quantum physics you know that our world is truly random.
00:28:11 Speaker 1
Yeah, so that we have randomness as inherent quality of our universe, although I somehow followed.
00:28:20 Speaker 1
Their ulstein's intuitions say that.
00:28:24 Speaker 1
God cannot throw the dice.
00:28:27 Speaker 1
And uh, on the other hand, I also think that disconnectivity systems, they may inherently have something there that makes it super hard, if not in principle, being not capable to be explained in in a, in a, in a human understandable.
00:28:48 Speaker 1
And you know it.
00:28:49 Speaker 1
It it could be again, that this state is never being achieved with this 100% predictions because there could be black boxes that cannot be in principle be resolved.
00:29:02 Speaker 1
Like connectivity systems and the randomness that our universe potentially has.
00:29:09 Speaker 1
So we've talked now about values and facts, and I would like to talk about another quality of knowledge.
00:29:17 Speaker 1
Since you're focusing on knowledge on a social level, there must be also.
00:29:22 Speaker 1
An individual level.
00:29:24 Speaker 1
And maybe you could outline the differences in between these sorts of knowledge, meaning the social and the individual one.
00:29:33 Speaker 2
So the mainly and and and really a lot learn in a social context, but we still also learn as individuals and we as individuals respond differently to these social stimuli that we are getting from our environment.
00:29:49 Speaker 2
But just think of us as children.
00:29:51 Speaker 2
We were all learning from our parents.
00:29:53 Speaker 2
1st and then started developing.
00:29:58 Speaker 2
But when we go kind of further the the the there is in my eyes a big difference which we touched only briefly.
00:30:08 Speaker 2
That's this what is rational for an individual and the pursuit of knowledge is not the same as what is rational for the group in the pursuit of knowledge so.
00:30:18 Speaker 2
For an individual, it's often very beneficial to just imitate and learn by imitating, or follow the group and the majority in the group.
00:30:30 Speaker 2
In this process of learning, that will probably Max.
00:30:33 Speaker 2
The knowledge of the individual, however, when we think about it on the social level, we might want these diverse approaches between different people and so think of a scientist who works in this specific environment, and there I always like to connect both the social aspects.
00:30:53 Speaker 2
And the epistemic aspects.
00:30:55 Speaker 2
So we have now scientists that are really socially struggling with early career researchers on temporary contract.
00:31:04 Speaker 2
They need to publish where they perish and for them it's really rational to publish in the field, which is mainstream field or, but for the scientific community as a whole, there is this benefit that they explore some neighbors idea and there.
00:31:23 Speaker 2
A lot of different phenomena with respect to social epistemology occur.
00:31:28 Speaker 2
One of them is this idea of the wisdom of the crowd, which means that we should in order that it occurs, we have to explore different approaches and different ideas independently.
00:31:39 Speaker 2
Of each other.
00:31:42 Speaker 2
To maybe introduce what it means for crowd to be wise.
00:31:48 Speaker 2
One of the examples is people who were supposed to estimate the date of an ox on the market, and it turned out that the.
00:32:00 Speaker 2
Majority estimate was better than an estimate of an expert.
00:32:05 Speaker 2
Of the weight.
00:32:07 Speaker 2
There we have this first contradiction retention and so it's more like attention that we assume.
00:32:14 Speaker 2
For instance, when people are running that the person who runs fastest has the higher running speed than the average runner of people.
00:32:24 Speaker 2
Of that group.
00:32:25 Speaker 2
But in science, we assume that the group always knows more than an individual expert.
00:32:32 Speaker 2
And the aim in social epistemology to maximize this knowledge on the group level and to see how we can then kind of by crowdsourcing knowledge, discover the best, the most accurate.
00:32:53 Speaker 2
Facts with in opposition to.
00:32:57 Speaker 2
Individual expert opinion.
00:33:01 Speaker 2
Some other examples often used in the literature was how many entries and how surprisingly accurate entries on Wikipedia are with respect to the in comparison to Britannica, which is considered to be kind of one of the most influential.
00:33:17 Speaker 2
Encyclopedias in the world that has this.
00:33:20 Speaker 2
Word review.
00:33:22 Speaker 2
While uh, Wikipedia is rather open and in a sense democratic and perform surprisingly well and and and this potential of the group is of course not always there because we have other problems in gaining knowledge than individuals.
00:33:43 Speaker 2
Then the group performs all the sudden, where as an individual, because for instance they end up in an information cascade.
00:33:51 Speaker 2
What would that be?
00:33:52 Speaker 2
That would be just following or imitating another person without the real knowledge behind.
00:33:59 Speaker 2
An example is when, for instance a a price of a book is supposed to be formed that actually allegedly held that happened or be in relationship.
00:34:11 Speaker 2
We learned that it.
00:34:12 Speaker 2
Happened one bookstore by accident.
00:34:15 Speaker 2
Online pay posted a very high price of a book and another book.
00:34:19 Speaker 2
So just.
00:34:20 Speaker 2
Copied this price but no one was in reality buying the.
00:34:24 Speaker 2
So they were just following each other, but there was no no real information behind it.
00:34:30 Speaker 2
And this tendency of information cascades.
00:34:33 Speaker 2
We see very frequently.
00:34:35 Speaker 2
For instance, we see it when people are sitting in one restaurant and across the street in the exactly same restaurant.
00:34:43 Speaker 2
There is no one sitting.
00:34:45 Speaker 2
Because again, there there is something irrational behind it, but we also see it for in even in animals.
00:34:52 Speaker 2
So in ants that just follow one after the other, their leader and search for food, but might even in the end die because they cannot find.
00:35:01 Speaker 2
That food source and this very kind of very deeply in our nature, so to to, to be to to, to express such behaviors.
00:35:15 Speaker 2
So to follow and imitate, and we should always, whenever we think about this cognitive behaviors, they always have.
00:35:22 Speaker 2
Certain benefits, because we learn by imitating very successfully.
00:35:26 Speaker 2
But sometimes it's not optimal and another maybe interesting point.
00:35:31 Speaker 2
So that's why we want this diversity.
00:35:34 Speaker 2
Diversity should help us against the information cascades because it can break the information cascade.
00:35:39 Speaker 2
It can give us fresh and different perspective, but it's also always again important and maybe that relates to your.
00:35:46 Speaker 2
Previous comment that kind of this diversity has to have certain limits, so this diversity in this context has.
00:35:52
To be meaningful.
00:35:55 Speaker 2
And then we're just think about this example with the with the the estimating the weight of the.
00:36:01 Speaker 2
Ox, if I would be supposed to estimate someones weight in kilograms I could do it, but if I would be supposed to estimate someone's weight in pounds that would be very hard for me.
00:36:11 Speaker 2
Because I don't.
00:36:12 Speaker 2
Know what's the ratio?
00:36:14 Speaker 2
£15 in kilograms. I could easily learn.
00:36:16 Speaker 2
But and then someone who wouldn't have these concepts, of course, it's then this is just a random guess and cannot no knowledge can be extrapolated from it.
00:36:29 Speaker 2
So the assumption behind this wisdom of the crowd is that we do have some diverse ideas, but these ideas relate to.
00:36:36 Speaker 2
The reality that we have some.
00:36:39 Speaker 2
Truthfulness in our beliefs so that we have some degrees of probability that we are right or that we at least understand concepts around us.
00:36:47 Speaker 2
If we are just asked about to predict results of presidential elections in some country that we don't, we are not familiar with.
00:36:55 Speaker 2
Of course, we cannot give any insightful information about it.
00:36:59 Speaker 2
And the second assumption there is the law of big numbers.
00:37:04 Speaker 2
So the wisdom of crowds occurs when we have the bigger the crowd with independent information.
00:37:13 Speaker 2
We are more likely to to to reach some accurate knowledge, which will then surprisingly maybe be or not maybe.
00:37:23 Speaker 2
But that's the assumption we higher than the knowledge or philosophy gets at the moment.
00:37:29 Speaker 1
Super nice many interesting topics to maybe discuss in future with you again.
00:37:35 Speaker 1
Plus it was a great pleasure to talk with you today.
00:37:38 Speaker 1
It was really fun.
00:37:40 Speaker 1
And yeah, thanks a lot.
00:37:41 Speaker 1
And yeah, hopefully see you again soon.