QI2016 Keynote speaker: Bas van Fraassen

so pleasure to have you guys here many familiar faces have wonderful speakers keynote speakers and widest speakers and most relaxed Mesa presenter so we're going to actually be filming that so we start today with boffin pradhans BOTS a distinguished professor of law at San Francisco State came to us after a fee for widening and grace what I didn't hear bus is of course he's also I'd like to welcome boss so well which is in 97 now what exactly was different organs outside the system States on the basis of observations so estate is assigned to the system being studied no parameters now you can study the study system is so it really doesn't make any sense at all but the measure themselves are just interaction I'm sure you all know now there's an intention because in fact reconciling B and C that's what's called the measurement problem so all the different versions of the all these different interpretations are being proposed and each of them offer the solution so so the weather the action there are new approaches they parked back to the external observation - in some respects to some extent inspired by this pathology because it's a mistake so that's the end of the the outcomes that we have registered in our measurements what are the probabilities or the various outcomes of the next a black box process as far as what we can see at the same time we have a model set up so that the state of the system changes it's not going to be a single mistake so how does the business reacted to these records well clean the wall says honey different states and also depending on the information about the setup that you start reading a ton of different states so what the states represents is not what the system is like what the state represents is the information that we have so in this situation but if the story is or n equals so the mechanical story that's like I think that their interpretation of in this family I want to say something more specifically it's the mother Valley it's information theory oriented to some extent in some important respects includes back to the external observation view at the same time he tries to dissolve the kind of pension let people away from that so in rebellious world just like monoi Mandi insist that an observable hasn't value if nobody if the state is an eigenstate so and if the states are yeah observers in observers so he had he knows something about the state of the observer a Travis system he has a different state or not I'm sure you're already thinking about of course I can try to answer questions okay so so I try to say bye I think that the extra motivation you in one form or another is something very it's something we should take seriously is constantly in the deterministic process well in formal epistemology we are talking about ages it states of opinion information States in the same way is this process the deterministic process has been until being interrupted right it's a transformation of economic group well what are the problems in any case homology we're dealing with knowledge belief an opinion informal in this technology it's basically up in that video representative right and a person has probabilities for what is going to happen then you get some new input we call it experience difficult experience but you know it could be reading my opinion change because so good but it is from perception or reading or whatever input signal strain or what's your next but still get opinion indeed but then equal strings of course is focus going to leave things open and the question is which posterior opinion that satisfy the constraint is the one that should go for that's the form of the problem now some results are so well now there is a second development of this sort was Richard Jeffery proposed that sometimes he comes through but you suddenly realize that you need a new probability I guess the easiest example would be I think it's open the rainy day and then in the morning I gave the better forecast 80% chance of rain ok don't become certain of the accommodation but change it to probability 8% and his rule also quite straightforward he says divide the possibilities probability this part that part again this is very similar for you very soon for a sympathy or argument there'll be more complicated right that says Oh posterior is supposed to be a little bit function of your prior to echo saying so what if I only need a part of the weather forecast and the forecaster says the probability of rain and so for today will be equal or suppose that I've he can do odds they're also greatest abilities are just about the office okay or it's a stage of conditional probability how do you solve those problems well physicists interchange any general form that takes that really covers all those old ones I just mentioned you've got a quantity with an expectation value and the input is low mid sized much so definition of expectation value it's so changed changed ever suggested and it was using information and what he proposed is so to the prior so these are the information that comes from the ratios the odds so you never see the probabilities first you see the Oscars that's not yes so a probability factor like on 541 to about an off spectre-5 speak two and six for why am i doing this because it was perfect the set of all practices if someone figured we didn't space and so if those ratios distance then let's say now now what we're looking at is we're looking at respect to space certain region small region as the probability factors a larger factors well first of all the quotient that's supposed to be important and secondly linearity because so asymmetry will be a transformation that preserves just that it preserves the quotients internet preserves deep in the air but that's important because that means that we can not think of it as the kind of transformation matter what your priors well maybe system it's subject to change and sustained well here's the analogy then a uniform motion they'll be similar V and one parameter group of uniform destinations which is far far away from the prior right what people to say is physics biology economics and so simply said now that the simplest example is a person who is making investments so here's my goal at age 65 million dollars I could be fired but I have emergencies how do I teach my goal by 865 so I have infinity lower interest rate should I say it satisfies equals 20 quid on while minimizing the move away from your prior a general is by the elements the components they are all deformed II case frz Percy at the time right so he survived they want it to be these are values numbers the posterior expectation value that quality should be is that's like saying I'm fighting interest rates and then we have a theory so first devastating my posterior it has to be probability function secondly as genes required I have a new input which is the expectation value for a given quality Q well in that case that exists a single number a constant W such that the transformation of the prior to the posterior is fine a uniform motion the basic components very simple it's just easy every component it's raised to the power the same and I did this very fast no it looks pretty upset right it's no no it's not for your transport but it's yeah it has the same slope form yeah right so okay that's one thing to say about any further mathematician you're right but here's what I see that the meaning of the fear gives exactly the same answers in applications as Jesus so that's what I wanted I will interfere you know retains had only a suggestion but theorem I have sympathy with development which tells you that's the analogy and say interrupted by the inputs in the different than the sender version of formerly Stevie it's the same kind of formalism the same kind of reasoning and it just I could only see the reasoning coming by 35:27 is because before that this question which sounded rather something historical question let's say just for sake of argument the three hundred years ago they already had quantum mechanics so so you've got lots of scientists like George Berkeley and a car alignments and could have had no problem say oh we need an external observer god fine yeah scientists agree yeah so all of those problems wouldn't wouldn't really clean but the determinism problem well hang on if there's this external observer but we all the great garden and people's find that but we still can't be you know the observers are inside the cross and himself human being is absurd he can do something to stay to the rest of the world so how can he say that we've you know gotten away from the problems well your addition describes the world as a whole that as an observer himself would never have the knowledge of what that state but when the industry rule is a whole you can nevertheless say oh but you know this in the 17th century would not have made sense so I mean it's determinism and everybody said so we could effectively put the in determinism we see tend to human frailty is very interesting to me I'm a nature follower of the interpretations of quantum mechanics on the Wikipedia site where unambiguously my belly is playing for the wave function to be not real so that does button up with the idealists yeah but is that because it's information because all other respects he's so close to where where it's real so how much information but the very function is or is the word we have to define but what will meals that's what most I want to say you actually say I mean I wanted to explore this I want to see what theorem was possible but you know that's outside this obviously because we can't argue with Einstein that you don't need something like consciousness or special relativity but I don't see how that applies to the measurement problem bigness trend and notwithstanding and if you if the machine and tangles with a quantum system that doesn't necessarily mean the results are going to collapse right so my question is what's your opinion of the Penrose which actually you know what's the observer some machine but that wouldn't really cause the result superposition in the world the states are collapsing but okay because I said I think otherwise of course when you function so the symmetry prior something which is this is technical issues far generalize the well the first problem it has to be so what do you do with zero so if you think about so what do you do right now just different ways of sorry with that I mean then so yeah I think like because st. region should never go back I we have a very long so what I was proposing is perhaps and satisfied following the program we could say a shortened coffee break and start at 10:30 instead of 10:45

Channel: Jose Acacio de Barros

Views: 706

Rating: 5 out of 5

Keywords: epistemology, QI2016, Bas van Fraassen

Id: hGHkmOxtJBA

Length: 60min 13sec (3613 seconds)

Published: Mon Jul 25 2016