COLLOQUIUM: 'The Physics of Can and Can’t': from the universal computer to the universal constructor

Ғылым және технология

06 February 2020
Speaker: Chiara Marletto, University of Oxford
Abstract:
The theory of the universal quantum computer has brought us rapid technologicaldevelopments, together with remarkable improvements in how we understandquantum theory. There are, however, reasons to believe that quantum theory mayultimately have to be modified into a new theory: for instance, it will have tobe merged with general relativity, to incorporate gravity; and some claim thatit may be impossible to have quantum effects beyond a certain macroscopicscale. So what lies ahead of quantum theory, and of the universal quantumcomputer? To shed some light into these questions, we need a shift of logic inthe way things are explained. Specifically, one can adopt the approach wherethe basic assumptions are general principles about possible/impossibletransformations, rather than dynamical laws and initial conditions. Thisapproach is called constructor theory. I will describe its application to a handfulof interconnected problems, within information theory, thermodynamics, and evenquantum gravity. This ‘Physics of Can and Can’t' may be the first step towardsthe ultimate generalisation of the universal quantum computer, which vonNeumann called the 'universal constructor’.

Пікірлер: 94

  • @destructorextreme707
    @destructorextreme7073 жыл бұрын

    To Chiara Marletto > "if the universe is a single substance which self differentiates; then the principles which allow that behavior would yield the true grand theory of construction. From this we could design technology which alters space time as we see it.

  • @Evan2718281828
    @Evan27182818284 жыл бұрын

    41:00 is eta the angle the system has gone around the circle of reservoir qubits?

  • @Evan2718281828
    @Evan27182818284 жыл бұрын

    53:00 is cool, asking what happens to the born rule if you vary constructor-theoretic principles of information

  • @Nightshift10000
    @Nightshift100002 жыл бұрын

    Is a universal constructor (like in Deus Ex) through nanotechnology even possible to make & if so how close do you think we are to making one?

  • @alvarofernandez5118
    @alvarofernandez51182 жыл бұрын

    This reminds me of the Buckingham Pi theorem, which you may use to reduce the set of possible dynamic *equations* which may result from an experiment, by requiring dimensional consistency. With constrctor theory, we are requiring that infinite precision be possible; else the candidate dynamic law is impossible. Hopefully im not completely off base...

  • @KaliFissure
    @KaliFissure3 жыл бұрын

    Stephen @Wolfram has created exactly this. A way of developing a universal Turing machine so that one could clip away rulial (dimension of computational rules) space until we find our universe by constraint. All conservation though comes from a simple principle. The universe is of a piece. It is a single unit and so only internal form can be modified

  • @rv706
    @rv7063 жыл бұрын

    I have a hunch that this theory will have more interesting applications to philosophy/epistemology than specifically to physics

  • @peaku8129

    @peaku8129

    3 жыл бұрын

    Hope that's not the case. Radical ideas are always met with skepticism that lacks argument.

  • @rv706

    @rv706

    3 жыл бұрын

    @@peaku8129: I'd be happy to be proven wrong, but I'm skeptical of grandiose claims, such as the ones by the authors involved in that theory, especially if they're accompanied by a lot of words instead of a lot of peer reviewed papers and clearly stated results. But never say never...

  • @greycookie666

    @greycookie666

    3 жыл бұрын

    @@rv706 it is theoretical physics at it's most theoretical. I am curious, what sort of clearly stated results from papers do you expect?

  • @rv706

    @rv706

    3 жыл бұрын

    @@greycookie666: Idk, is it at least possible to rederive quantum mechanics from constructor theory? If yes, is this done seriously or is it just handwaving? - Does CT make any testable predictions? --- I'm not familiar with CT, but in words it is said to obtain "everything" (quantum mechanics, thermodynamics, ...) and I am skeptical of things that are claimed to do everything.

  • @peaku8129

    @peaku8129

    3 жыл бұрын

    @@rv706 nobody knows for sure where this theory is going. I guess the right attitude is to be an skeptic-optimist: Take every argument with a grain of salt, but be positive towards its development.

  • @Evan2718281828
    @Evan27182818284 жыл бұрын

    What parts of quantum theory do superinformation media not have to obey? When I read the CT Info paper they were able to derive a lot of quantum properties from their defn of superinformation.

  • @kamilziemian995
    @kamilziemian995 Жыл бұрын

    Interesting talk.

  • @zephirinjeff
    @zephirinjeff3 жыл бұрын

    It is possible to understand, the beginning of human existence self of consciousness; if, we decided to go back at the primordial of the first iota internet user.

  • @GrahamBessellieu
    @GrahamBessellieu4 жыл бұрын

    Universal Constructor: an object that has within its repertoire all physically possible transformations (not just computations).

  • @koltonjase2284

    @koltonjase2284

    2 жыл бұрын

    You all prolly dont care at all but does any of you know of a way to get back into an instagram account?? I somehow forgot my account password. I appreciate any tricks you can give me!

  • @jamiegenesis7342

    @jamiegenesis7342

    2 жыл бұрын

    @Kolton Jase Instablaster =)

  • @koltonjase2284

    @koltonjase2284

    2 жыл бұрын

    @Jamie Genesis thanks so much for your reply. I found the site thru google and Im trying it out now. Takes quite some time so I will reply here later when my account password hopefully is recovered.

  • @koltonjase2284

    @koltonjase2284

    2 жыл бұрын

    @Jamie Genesis it worked and I finally got access to my account again. I'm so happy:D Thanks so much you really help me out!

  • @jamiegenesis7342

    @jamiegenesis7342

    2 жыл бұрын

    @Kolton Jase glad I could help xD

  • @araneascience9607
    @araneascience96072 жыл бұрын

    I think that the QUNIVAC paper from David Ritz Filkestein is very similar to the focus of this theory, would be great if someone merge both of them.

  • @KuldeepSingh-xy9ji
    @KuldeepSingh-xy9ji3 жыл бұрын

    nice

  • @N05Y573M
    @N05Y573M2 жыл бұрын

    I love that woman... ❤️❤️❤️ Multidimensionaly!!🌹

  • @johnnywilliams2641
    @johnnywilliams26412 жыл бұрын

    Im pretty sure Wolframs new science gave me a different idea of what a universal computer was

  • @warrenberkholtz8410
    @warrenberkholtz84103 жыл бұрын

    Can someone smarter than I on modern physics, comment: - would this Constructor Theory correlate to better modeling of matters related to Entanglement, Network Effects and Metcalf's Law ?

  • @usaintltrade
    @usaintltrade Жыл бұрын

    🌹

  • @waltertanner7982
    @waltertanner79823 жыл бұрын

    Should this universal computer be an implementation of a Turing Machine?

  • @david203

    @david203

    3 жыл бұрын

    I think their point is that a Turing Machine is insufficient for predicting the world around us, in particular the rich area that lies between very simple systems and very complex or chaotic systems. I'm not sure.

  • @waltertanner7982

    @waltertanner7982

    3 жыл бұрын

    @@david203 The question is: will it in principal be programmable (by whom) or will it just „somehow“ (BlackBox) gather data and TheSolution pops up after some time (42? Who will understand/believe/use such a solution? Is it neccessary that someone/something u/b/u it?)

  • @david203

    @david203

    3 жыл бұрын

    @@waltertanner7982 Well, these are excellent questions. Unfortunately, the videos on Constructor Theory are so long-winded and vague that we have no idea what the answers are. I guess the founders (David Deutsch, et.al.) want to transform physics from a random collection of experimental and theoretical knowledge to, well, something else, I guess. See en.wikipedia.org/wiki/Constructor_theory, which speaks of "abstract specifications of transformations in terms of input/output pairs of attributes," whatever that means. All I can say is that if it can predict the way a creature looks and acts based on its DNA, or, even better, if it can design DNA to create a creature that looks and acts according to a specification, then it is most certainly successful in answering questions that physics is not even ready to ask. We shall see.

  • @uazuazu

    @uazuazu

    3 жыл бұрын

    She specifically says in the lecture that a Turing Machine is not a universal computer, because it cannot create a copy of itself. So a Turing Machine cannot do what a living cell does. So it is not universal.

  • @david203

    @david203

    3 жыл бұрын

    @@uazuazu I certainly agree. Turing Machines have to do only, ONLY with the computation of algorithms, step by step operations that end by halting. Nothing else. While computation is interesting and useful, it is not the whole operation of nature, or the whole province of physics. I wish the Constructor Theorists great luck, which they will need since they offer no actual construction that duplicates nature in any new way. Correct me, someone, if I'm wrong, but it all seems to be just words. After ten years of preparation, it's still just words. I want to know which cellular automata generate chaotic behavior, just by inspecting the automata rules. I want to know how to construct DNA that will build a given structure made up of proteins. Isn't this the province of Constructor Theory? And if the Constructor theorists are so innovative, how come they don't embrace the de Broglie-Bohm interpretation of quantum mechanics, which has been validated by experiment and solved QM's basic problems and eliminated its mysteries back in 1952?

  • @martinbiehl4596
    @martinbiehl45964 жыл бұрын

    18:41 How do I know whether a theory allows for a definition of an energy? What is the criterion? What is the general notion of energy used here? Is there an example of an energy definition in system without specified dynamics?

  • @martinbiehl4596

    @martinbiehl4596

    4 жыл бұрын

    Hm, maybe "conservation of energy" is just a name for the law of physics that "perpetual motion machines are impossible". What perpetual motion is, is also not really clear to me but I guess it can be more easily defined using transformations of substrates than energy.

  • @MS-il3ht

    @MS-il3ht

    3 жыл бұрын

    Is it not like an engine of the possible?

  • @efraimcardona8452

    @efraimcardona8452

    3 жыл бұрын

    It is a time evolution conserved quantity of a system whose differential is proportional to the work.

  • @sam71119
    @sam711192 жыл бұрын

    When we are constrained via "known" physics and Quantum etc, how can anything new be uncovered... ? How about stepping beyond the known... then work out why it works the way it does...? Often the narrative creates the result... so drop the narrative and allow dreams in the still of the night to foster new ideas... (just saying)…

  • @audiodead7302
    @audiodead73023 жыл бұрын

    Not sure I am fully understanding what Chiara is proposing. I can sort of see how you might try to predict the limits of the future development of a technology using it. e.g. travelling faster than the speed of light is impossible. Therefore interstellar travel is a non-starter (alpha centauri if 4 light years away). I wonder if one can use it to answer questions like: Is it possible to simulate the entire universe inside a (universal quantum) computer? Is that what this is about?

  • @gaminghunt5837

    @gaminghunt5837

    3 жыл бұрын

    How can we do that?that is a huge information.

  • @audiodead7302

    @audiodead7302

    3 жыл бұрын

    @@gaminghunt5837 Agreed. But there is a lot of speculation about how powerful future generations of quantum computers could be. If the 'many worlds' interpretation of QD turned out to be correct (big if), quantum computers might effectively be limitless (i.e. Infinite parallel processing).

  • @gaminghunt5837

    @gaminghunt5837

    3 жыл бұрын

    @@audiodead7302 but how will we connect the computers together?what will we do if the computers malfunction?Chaos is everywhere.

  • @sunroad7228
    @sunroad72283 жыл бұрын

    "No energy system can produce sum useful energy in excess of the total energy put into constructing it. This universal truth applies to all energy systems [including what's called the Universal Constructor]. Energy, like time, flows from past to future".

  • @En_theo
    @En_theo3 жыл бұрын

    @38:40 Why does she consider that the machine Y can't cool down by using the energy of the weight ? Just use a pump that will decompress a gas and it will cool down. I'm not sure I got her point there.

  • @uazuazu

    @uazuazu

    3 жыл бұрын

    Where is the energy going in that case? It's going to the environment. The machine in one direction is turning potential energy into heat energy. It cannot go in the reverse direction and turn heat energy into potential energy. What you're talking about is a completely different "reverse" transformation which isn't a reverse transformation at all: where potential energy is used to move heat energy from the container to the environment.

  • @En_theo

    @En_theo

    3 жыл бұрын

    @@uazuazu What if there is no atmosphere around the machine that I described ? Pulling the cylinder of the piston will increase the volume of the gas under the piston. That will cool the gas and there is no heat generated, quite the opposite : the gas will absorb the heat of the surrounding parts and the total amount of heat will be smaller. The machine X she described has been specifically chosen so it cannot reverse heat, but that does not prove it's impossible. It's not a fundamental law.

  • @uazuazu

    @uazuazu

    3 жыл бұрын

    @@En_theo Still, the machine you describe is not the reverse of the original machine. It does not take the end-state back to the starting state.

  • @En_theo

    @En_theo

    3 жыл бұрын

    @@uazuazu Actually it does, by compressing the air (the compressed air will have an elastic reaction) and the system will. But that's not what she explained, she compared the fact that stirring backwards could not cool the system and don't forget that system X and Y are two devices, not the same device in different states. As I said, she chose a machine (stirring) specifically so that it cannot be reversed. But that is meaningless, since her point is that reversibility was universally "impossible" while it's just impossible with "that" machine. So I still don't see the point here, maybe she forgot to explain a part that was obvious to her but not to the public.

  • @hanniffydinn6019

    @hanniffydinn6019

    10 ай бұрын

    A perpetual energy machine is impossible, that’s the “constructer” rule! Ask yourself Why is that? What physical rules apply? 🤯🤯🤯🤯🤯

  • @johnnywilliams2641
    @johnnywilliams26412 жыл бұрын

    Fucking love Chiara! Grazie bella (:

  • @timemechanicone
    @timemechanicone2 жыл бұрын

    Sequence ⭐️+1 bit - to use natural atoms around us to create - anything the rules allow! 1bit of dense information - structural time. Using time mechanics & mathematics! ♾4D depth not horizontal Each information structure is a independent technology. Elemental Algorithm A sequence of mathematical interactions recreated via Fibonacci, prime mn ++ sets = preload last bit data! Braided create a unique sequence through infinity information to collect pre entanglement information. Universe / persons - choices = Updates 1bit - @ each infinite moment! Same 1 bit holds compressed time information - all previous to current information no need for previous loads! All can be extrapolated from 1bit. Each moment is 1bit updating No space - depth in time! The sequence expands & retracts! P N P for children of Ai ? Increasing information thus entropy! Decay! Raw atomic information turned into - Organic built - technologies? You just need the code! The or a specific- Sequence. Zero else.. Organic computer🤓🖖♾ Geometric information structures Geometric moments Geometric timeline Geometric Time-lines Geometric universes! INTERCHANGEABLE Paths in information! TO? 🌬⭐️ mergeability ..not a word is now for a moment. Priceless information- screenshot have it checked. Come curse me if wrong! 😊😌🖖 welcome to DMs are open.

  • @johnnywilliams2641
    @johnnywilliams26412 жыл бұрын

    And what the hell is an electron made of? lol Is planks constant space itself? An electron changing energy levels shoots a tiny bit of space

  • @johnnywilliams2641
    @johnnywilliams26412 жыл бұрын

    Raising money for science is easy. Step 1: Get a company started and Introduce it to the American stock market. Step 2: Make a shady deal with Elon Musk to tweet about your stock. Step 3: Sell all of your shares before Elon sell his and then repeat steps 1 through 3. It's called A PERPERTUAL MONEY MACHINE and it does exist.

  • @gxfprtorius4815

    @gxfprtorius4815

    2 жыл бұрын

    Nice. It only works as long as other peoples' money drives it though, so it's the same thing, isn't it?

  • @glcpit7797
    @glcpit77977 ай бұрын

    smoke in the air ...

  • @waltdill927
    @waltdill9273 жыл бұрын

    Need more math and less vague outline. Also, physics has "painted" itself into a very difficult corner lately, when you consider the elusive nature of black energy/matter. These ideas are useful philosophy, but the epistemological "arguments" are broad-brush stuff.

  • @hinteregions

    @hinteregions

    3 жыл бұрын

    If it had more maths then it would not be a generalist lecture for interested laymen but a specialist one for qualified people. Did you simply not listen to her introductory remarks or were you at the time too busy working on your comment? If you are just so much better than she is at giving these sorts of lectures where is your channel, never mind your PhD XD

  • @uazuazu

    @uazuazu

    3 жыл бұрын

    The whole point is to see what you can derive without getting into the details. It seems that they have already derived some useful results, e.g. predicting that if a gravitational field can be used to entangle quantum bits, then gravity must be non-classical (at least partly). So if that gravity experiment can be done (or observed in the cosmos somehow) and other people agree with the reasoning, then that takes physics one step forward.

  • @waltdill927
    @waltdill9273 жыл бұрын

    BTW: We already have a universal computer; it's called "consciousness of the conscious other" ---

  • @climatebabes
    @climatebabes3 жыл бұрын

    Unversal computers can do any computation. They don't claim to do be able to do more, like replicate for instance.. So I don't quite follow your argument. You should have a solid one imho.

  • @timemechanic5055
    @timemechanic50553 жыл бұрын

    Physicists have a very poor understanding of probability theory compared to, say engineers. This proposal takes them back even further. Replace Born's rule. "Derive the Born-rule-like expectations for repeated measurements" what does this word-salad mean? Are you proposing going back to frequentist outlook! This is just playing around with semantics and ignoring the correct way to think about probability. John von Neumann would be rolling in his grave.

  • @rexdalit3504

    @rexdalit3504

    3 жыл бұрын

    TM, I find it odd that you should say this about engineers & physicists... notions like Fermat distributions, delta functions and their derivatives & integrals, orthogonal transformations of density functions, imaginary density functions (i.e. probability amplitudes), probabilities as solutions of integro-differential equations, all statistical mechanics, all thermodynamics, and on and on and on... were all developed by physicists decades (or longer) before engineers used them. (Your use of the term 'word salad' to describe a probability procedure that is well established in physics should be a warning sign to you. BTW, QM now has at least 9 diffeomorphically homologous representations that can produce calculations homomorphic to the Born rules.) My general rule-of-thumb used to be that physicists were about 50 years ahead of engineers in matters of theory, and I suppose this is still more or less the case. The good news is that engineers have a huge refreshing pool of futuristic ideas from which to quench their thirst. Drink deeply and enjoy.

  • @timemechanic5055

    @timemechanic5055

    3 жыл бұрын

    @@rexdalit3504 Let me explain! Physicists did indeed develop statistical mechanics at an early stage - but critically this was before Kolmogorov's formulation of probability theory in the 1930's and the subsequent development of stochastic processes. Instead they continued to use primitive concepts such as "ensembles". It is quite telling that your list of mathematical tools are mostly 19th century (did you just make up Fermat distributions?): measure theory and abstract integration don't make your list (essential for modern probability theory), nor W*-algebras which is the noncommutative analogue. Engineers adopted the correct mathematical approach to describing noise and extracting information and subsequently controlling systems from the 1960's onwards as can be seen in the work of Kalman, Kushner, Stratonovich, Zakai, etc. Are you familiar with these? The quantum extension was worked out in the 1980's, but physicists still mess around with "collapse of the wave-function". The primitive conceptualization of probability used by physicists is problematic and is wholly inadequate for formulating and addressing practical problems solved by engineers - and this extends to quantum to. Talking about word-salads, can you give any of the "diffeomorphically homologous representations that can produce calculations homomorphic to the Born rules"?

  • @charki40
    @charki403 жыл бұрын

    This is interesting. Just for skeptics here. Can you tell me the quantum explanation of thought in the human mind, or what is conciseness. Or how a proton can be a particle then a wave through the slit experiment...why and what is it happening in those states, or how entanglement actually works beyond an observed description. Is it not some form of information exchange of sorts? Let us see where this may led us. Stop being such closed socially biased circuits to what is potential knowledge.

  • @johnnywilliams2641
    @johnnywilliams26412 жыл бұрын

    I still feel like there has to be a way to use capillary action to get a drop of water to drip. What about 20 or 30 tubes that produce capillary action and they come together to make a droplet large enough to break the surface tension. Isn't the universe as a whole a perpetual motion machine or sorts. If we can get a droplet of water to fall we can make the worlds smallest water wheel and power a bugs house. I just feel like you scientist arn't trying. Then you got Veritasium with his stupid windmill bike that is obviously made up and doesn't work. Unsubscribe from his dumb channel.

  • @carloscastanheiro2933
    @carloscastanheiro29333 жыл бұрын

    I need the layman explanation. There's just so much I don't understand. I've never felt so stupid as today. It's the language you use. Please assume we're not all physicists and use words we can understand. We're not all familiar with the terms. Man I feel so stupid.

  • @KRYMauL

    @KRYMauL

    3 жыл бұрын

    I think it’s the shaper theory of Plato.

  • @carloscastanheiro2933

    @carloscastanheiro2933

    3 жыл бұрын

    @@KRYMauL You mean the Theory of Forms?

  • @KRYMauL

    @KRYMauL

    3 жыл бұрын

    @@carloscastanheiro2933 Yes, I think that’s essentially what this is.

  • @carloscastanheiro2933

    @carloscastanheiro2933

    3 жыл бұрын

    @@KRYMauL I don't know, I think that's just a tiny portion of it. She talks about Quamtum computers and it feels so complex to me.

  • @KRYMauL

    @KRYMauL

    3 жыл бұрын

    @@carloscastanheiro2933 That’s just an example.

  • @iainmackenzieUK
    @iainmackenzieUK2 жыл бұрын

    If she is willing to explore ways of raising funds for research , I am happy to donate 1000 euros for dinner with Chiara - :)

  • @madhavestark3173
    @madhavestark31733 жыл бұрын

    Is it only me who thinks this theory is non-fruitful philosophy ?

  • @findbridge1790
    @findbridge17903 жыл бұрын

    smart women are so hot

  • @hudatolah
    @hudatolah3 жыл бұрын

    You can present in your own language and have subtitles with translation rather than everyone around you pretending that your English is great. I can’t understand you. I see universities doing this stuff all the time. They put up a lecture with a professor that nobody can understand and then blame the student for getting bad grades in that class.

  • @JJviniciuss

    @JJviniciuss

    3 жыл бұрын

    I could easily understand her and English is not even my first language...

  • @hudatolah

    @hudatolah

    3 жыл бұрын

    @@JJviniciuss ok, and? I don’t see how your understanding helps me understand her. And why are you assuming English is my first language? It’s my fifth.

  • @JJviniciuss

    @JJviniciuss

    3 жыл бұрын

    @@hudatolah what I am saying is not that me understanding makes it easier for you. Is that you may not be fluent enough on the listening aspect of the language. And that would make sense since, as you said, English is your fifth language. Also, I never once assumed English was your first language... I never said that.

  • @AsifMehedi

    @AsifMehedi

    3 жыл бұрын

    @@hudatolah If you can't understand her English, which is clear and understandable, you probably shouldn't complain that it's other people's fault.

  • @BoggleMeBog
    @BoggleMeBog3 жыл бұрын

    Time makes things possible, but time is just causality, so things are possible because something else caused it, but that something else is infinite, so physics is doomed. This caused this, that caused that, this is this, and that is this, and forever in a loop. It's like an endless mindfuk of definitions and laws, all leading back to each other, all wrong, all incomprehensible. What is possible is what has happened, and what could happen, but anything can happen. Shit we're all here right? Which is basically impossible, but here we are. There is nothing impossible if it all is, and there is nothing possible if something isn't. It all works in a balance to where you gotta ask, where is that balance? Everything has a balance. Energy, mass, time, and that is the key. Somewhere things balance, but how can you say that something balances for a reason other than the fact of balance itself? What is the thing that balances, which is not balance itself? You get to a point where words just don't cut it. How the fuck can I define something that defines itself, or that is not definable?

Келесі