New Pi Formula (the extra physics bit) - Numberphile

แชร์
ฝัง
  • เผยแพร่เมื่อ 14 ม.ค. 2025

ความคิดเห็น • 56

  • @numberphile2
    @numberphile2  6 หลายเดือนก่อน +22

    The main video is at th-cam.com/video/nXexsSWrc1Q/w-d-xo.html and the interview with Sinha and Saha is at th-cam.com/video/2lvTjEZ-bbw/w-d-xo.html

    • @edernollivier
      @edernollivier 5 หลายเดือนก่อน

      Thank to you, care.

  • @MPS186282
    @MPS186282 5 หลายเดือนก่อน +22

    I've heard "is is" entering most people's vocabulary enough in recent years, but at 4:28 Tony pulls out a rare "is is is." Quite impressive!

    • @therabbithat
      @therabbithat 5 หลายเดือนก่อน +3

      "is is" is part of a cleft sentence.
      The thing about cats is...
      The reason why is..
      Where it is is...
      The thing is that some people have started saying it outside of this, or in cleft sentences that don't need it. "the thing is is"

    • @MPS186282
      @MPS186282 5 หลายเดือนก่อน +1

      @@therabbithat yes, and that sentence construction has gotten more popular recently to the point where, in almost every video or conversation, someone will eventually use "is is" as if it's a single word.

  • @Ishanaroya
    @Ishanaroya 6 หลายเดือนก่อน +32

    Great work! I love it when we get more details about these stories.

  • @Foxtr0t1337
    @Foxtr0t1337 6 หลายเดือนก่อน +109

    Pi and pions are related after all!🤣🤣🤣 It is hilarious that someone asked "How are pions related to the number pi" and got a no.

    • @georgerussell2947
      @georgerussell2947 5 หลายเดือนก่อน +7

      No backwards is on. Pi-on

    • @DrDeuteron
      @DrDeuteron 5 หลายเดือนก่อน +5

      well, there are approximatley pi types of pions, or exactly floor(pi).

  • @JoeKoenen
    @JoeKoenen 5 หลายเดือนก่อน +8

    I love watching Tony get excited about new papers & results. His joy is infectious 😁😁😁

  • @_notch
    @_notch 6 หลายเดือนก่อน +15

    I love numberphile so much

  • @iabervon
    @iabervon 5 หลายเดือนก่อน +14

    When you've just come up with a new way of doing a calculation, it's useful to put in some values where you know that the right answer is something well-known but not trivial. You can be pretty sure you haven't made a mistake if you find a new series that your result says will converge to pi, and taking a lot of terms gives a result that's very close to pi. If you come up with a way to calculate expressions that nobody know the answer to, it's hard to tell whether you have a typo along the way.

  • @joshhickman77
    @joshhickman77 6 หลายเดือนก่อน +22

    He makes a strong case the paper is interesting in a way totally unrelated to the pi thing haha

  • @ClassicPass_
    @ClassicPass_ 5 หลายเดือนก่อน +10

    Exciting things coming g from string theory.... in 10 years... lol🎉😂

  • @chicojcf
    @chicojcf 5 หลายเดือนก่อน +2

    Extraordinary visual output in this video re: CERN.

  • @peterpodgorski
    @peterpodgorski 5 หลายเดือนก่อน +9

    So they found a way to make the string theory equivalent of an effective field theory? Did I get that right?

    • @lepidoptera9337
      @lepidoptera9337 2 หลายเดือนก่อน

      Would it surprise you if string theory turns out to be an effective field theory? It would not surprise me.

  • @johnjeffreys6440
    @johnjeffreys6440 5 หลายเดือนก่อน +7

    Recipe for Pi, haha, I got that one.

  • @szkoclaw
    @szkoclaw 5 หลายเดือนก่อน +5

    Considering how complicated the formula is, does it give better approximation when computation cost is considered rather than just the number of steps?
    On another note, does it make string theory actually falsifiable?

  • @sherhy3689
    @sherhy3689 6 หลายเดือนก่อน +9

    Fun stuff

  • @ricardoabh3242
    @ricardoabh3242 5 หลายเดือนก่อน +4

    Could this imply that now some part of string theory can be tested?

    • @lepidoptera9337
      @lepidoptera9337 2 หลายเดือนก่อน +1

      If it allows them to calculate the low energy string spectrum, then yes. It does not necessarily lead to a practical result because of the string landscape. Unless it allows us to pick one specific version of the theory out of the 10^500 (or whatever) possible ones, the spectrum is still not uniquely determined. That problem will most likely not go away anymore than the equivalent problem in quantum field theory. Why is the low energy spectrum in QFT U(1)xSU(2)xSU(3) and not something else? We don't know. Well, actually, it's not even that if you are precise. You, as in your body, are a peak in the low energy spectrum of QFT. So is a water molecule, a goat, third avenue in New York and the star Beta Orionis. "Every possible thing" in physics is such a peak. We aren't calculating any of these solutions from first principles. We are happy to stop at the em field, electrons, muons, neutrinos, quarks etc., so we are basically cutting ourselves off at one massless and one solution that is a fraction of an eV (the neutrinos). And technically we can't calculate the locations of any of these peaks in QFT, either. We a simply fitting measurements to them. So it's not like that string theory can't so something that QFT can. They are pretty much equally "powerless" at the moment, it just gets misrepresented in the public discussion. (And no, I don't think string theory is the correct approach, either, I am just trying to qualify a little better what "theory" means in this sector.).

  • @frankharr9466
    @frankharr9466 5 หลายเดือนก่อน +1

    That's pretty neat.

  • @trwn87
    @trwn87 5 หลายเดือนก่อน +1

    So nice!

  • @JxH
    @JxH 5 หลายเดือนก่อน +5

    6:23 On-Screen (towards the bottom): *".....new....representations of....π, which show fast convergence."*
    In the 'Part 1' video, it was discussed how the convergence of this new formula was extremely slow, and the authors stated that they never claimed that it was fast.
    And yet I see, "...π....fast convergence."
    Is the π symbol quoted above the common 3.14..., or is it something else? (Just in case that's the explanation.)
    🙂🙂🙂🙂🙂🙂🙂 .: confused :. 🙂🙂🙂🙂🙂🙂🙂

    • @numberphile2
      @numberphile2  5 หลายเดือนก่อน +11

      I don’t think anyone said it was “slow”… the authors just said they were not claiming it was revolutionary.
      How fast or slowly it converges of course depends on lambda. High values of lambda of course do make it slow (Madhava speeds)
      But no matter what you set as lambda, it is not in the Chudnovsky ballpark.

    • @wwkk4964
      @wwkk4964 5 หลายเดือนก่อน +5

      The media sensationalized the claim and said it was extremely fast by modern standards. It wasn't but it doesn't mean the convergence is slow, it's still amazingly fast, just not like SOTA.

  • @inflivia
    @inflivia 5 หลายเดือนก่อน +1

    Could someone tell me a source for the graph found at 2:40?

    • @tinto278
      @tinto278 5 หลายเดือนก่อน +2

      Tanc 2D plot

    • @inflivia
      @inflivia 5 หลายเดือนก่อน

      @@tinto278 somewhat similar but not the same

  • @BarryMagrew
    @BarryMagrew 5 หลายเดือนก่อน +1

    Am I right that the Beta(1/2,1/2) gives not pi exactly, but (sqrt(pi)/2?

  • @tushar_bajpai
    @tushar_bajpai 4 หลายเดือนก่อน +1

    I found a new way too!
    [(cos²π + sin²π)-1+π] 😅

  • @kamilziemian995
    @kamilziemian995 5 หลายเดือนก่อน

    Did we learn something important about Euler function or not?

  • @tinto278
    @tinto278 5 หลายเดือนก่อน

    Tony uses a Mac and not a NeXT..😬

  • @untitled6391
    @untitled6391 5 หลายเดือนก่อน

    One question, why use this pi formula when we faster ones?

    • @nrml76
      @nrml76 5 หลายเดือนก่อน +2

      I think everyone in the video made it clear that this was just an interesting byproduct of their research and it was not published to offer a more efficient or accurate way of calculating pi. The importance of publishing these side notes is that someone in future might pick the baton up and develop it into something more useful. Most of the gamechanging scientific and engineering developments we now take for granted came about this way.

    • @untitled6391
      @untitled6391 5 หลายเดือนก่อน

      @@nrml76 got it, cool

  • @edernollivier
    @edernollivier 5 หลายเดือนก่อน

    Normally, it is strictly ruled.

  • @AJ5
    @AJ5 6 หลายเดือนก่อน +3

    Why is this unlisted??? I can't find it by searching

    • @numberphile2
      @numberphile2  6 หลายเดือนก่อน +19

      It will be publicly listed later.

  • @the_nuwarrior
    @the_nuwarrior 5 หลายเดือนก่อน

    Beta function regularization!

  • @davidwilkie9551
    @davidwilkie9551 5 หลายเดือนก่อน +1

    Enjoyed this more than a skeptical observer should?
    Euler's e-Pi-i 1-0-infinity superposition identification only needs a Geometer to rewrite the process of Scattering in Unit Circle modulo-geometrical Superspin-spiral Singularity-point interference, and everyone will be happier about why String Theoretical reasoning corresponds to log-antilog 2-ness i-reflection containment in 3-ness Cavity Resonance of Logarithmic Time Duration Timing Conception.

  • @theaveragepro1749
    @theaveragepro1749 6 หลายเดือนก่อน +4

    Hmmm B(1/2, 1/2) = PI? if B(x,y) is x!+y!/(x+y)! then (1/2!+1/2!)/(1!) is 1.77 which isn't pi. Don't understand that part.

    • @mgmanzanillo
      @mgmanzanillo 6 หลายเดือนก่อน +15

      The key thing here is that.
      Gamma(1/2) = sqrt(pi) and
      Gamma (1) = 1
      So, B(1/2, 1/2) = [gamma(1/2) * gamma(1/2)] / gamma(1/2+1/2)
      -> B(1/2, 1/2) = [sqrt(pi) * sqrt(pi)] / 1
      -> B(1/2, 1/2) = pi
      Hope this helps!

    • @MrAlRats
      @MrAlRats 6 หลายเดือนก่อน +6

      Gamma(n) = (n-1)!
      Gamma(1/2) = (-1/2)! = sqrt(pi)
      Gamma(1) = 0! = 1
      B(x,y) = Gamma(x)*Gamma(y)/Gamma(x+y)
      B(1/2,1/2) = Gamma(1/2)*Gamma(1/2)/Gamma(1) = sqrt(pi)*sqrt(pi)/1 = pi

    • @thstroyur
      @thstroyur 5 หลายเดือนก่อน +11

      The Gamma function isn't _identical_ with factorial (look it up on Wiki, it's a funny-looking integral); it just so happens to behave in the same way when its argument is a positive integer. This is related to what mathematicians call 'analytical continuation': the Gamma function 'continuates' (or extends) the factorial operation.

    • @888Xenon
      @888Xenon 5 หลายเดือนก่อน +4

      The Gammas on the top of the expression are multiplied, not summed.

    • @wesleydeng71
      @wesleydeng71 5 หลายเดือนก่อน +4

      You got the beta function wrong. B(x,y) = Γ(x)Γ(y)/Γ(x+y), and Γ(1/2)=√ π.

  • @Rakesh_kumar711
    @Rakesh_kumar711 5 หลายเดือนก่อน +2

    It's already discussed a year ago by Indian physicist 😂

  • @thesuccessfulone
    @thesuccessfulone 6 หลายเดือนก่อน +3

    That is an obscene formula.

  • @edernollivier
    @edernollivier 5 หลายเดือนก่อน

    Actually this is not right, because it has no mathematical fundations, with respect to the academic issue.

    • @lepidoptera9337
      @lepidoptera9337 2 หลายเดือนก่อน +1

      What does? String theory? What do you mean by "mathematical foundations"? That one can't give an axiomatization? Of course one can. At the end of the day all of these things flow back to Kolmogorov and partitions of unity. This one supposedly obeys general relativity in the low energy limit, whereas quantum field theory is only required to obey special relativity.

  • @RazvanMihaeanu
    @RazvanMihaeanu 5 หลายเดือนก่อน +2

    *Garbage!* There's only ONE accurate formula for Pi , and that's... *Viète's formula.*
    The reason is because it containts infinity, 1, 2, half of one... andf most importantly square root of 2.
    It has to do with the circle and the infinity of right triangles inscribed in the circle, where the hypotenuse is the diameter of the circle... and from this infinity of inscribed right triangles, there is only ONE isosceles triangle (if you use only half of the circle), with sides 1 and hypothenuse being square root of 2.

    • @DrDeuteron
      @DrDeuteron 5 หลายเดือนก่อน

      I guess as pi being defined as half the circumference of a unit circle, you need to address that fact in the formula. Anything else is number theory voodoo.