See brilliant.org/numberphile for Brilliant and get 20% off their premium service (episode sponsor) Order Tony's book Fantastic Numbers and Where to Find Them: A Cosmic Quest from Zero to Infinity Amazon US - amzn.to/3JYQbws - Amazon UK - amzn.to/3M3yvB8
@@namelastname4077 You can spend all your time contemplating the miseries of life and inevitablility of death if you want - personally I prefer to spend mine getting excited about fun cool things
Let's be clear about this. 1+2+3+... does not equal -1/12. The series is the result of a function definition that doesn't work at -1. However, it's true that there is another more complicated function definition that gives the same values where the first definition works, and also works at -1. It's that other function that has the value -1/12 at -1. A theoretical physicist tries to calculate something and gets the result 1+2+3+... . They guess that maybe they used the wrong maths, and maybe the right maths would give that other function so the answer is -1/12. If experiments then agree with this prediction the physicist becomes famous; if not they shrug and try a different way to calculate it. Edited : I typed +1 when I meant -1. Hey ho.
@@john_g_harris What 1+2+3+... equals, depends on your particular choice of how to assign values to infinite series. It's not possible to assign any finite value to it if you choose to adopt the standard definition but there are other definitions. The Ramanujan summation of 1+2+3+... does equal -1/12. Which particular definition is relevant to solving any particular problem can vary depending on the context in which the summation arises.
@@john_g_harris Regularization of the Riemann zeta function at s = -3 is used in calculating the Casimir effect and more generally in quantum mechanics there's a fair amount of renormalization where techniques are used to get a finite sum from a divergent series to get actual results. The argument that the sum of 1+2+3+ ... does not equal -1/12 because it uses a different method of getting the result comes up a lot. While it's important to recognize that yes, it doesn't mean "equals" in the same way as other "equals", this exact sort of thing has happened before. By the rules of basic arithmetic, the sum of a rational number and another rational number is a rational number. But take all the nonnegative integers and sum the reciprocal of their factorials and you get the transcendental number e. However, getting to this result, or for that matter getting the result of any convergent infinite series requires a different technique than basic arithmetic. This is not a controversial result today because people are used to the concept of limits and zero, but in the time of Pythagoras or Archimedes, it would have been jus as controversial as summing the positive integers to -1/12. There's an apocryphal story that a member of the Cult of Pythagoras came up with a proof that the square root of 2 is irrational and that the Pythagoreans were so incensed with the result because it broke the rules that they believed in that they took him out to sea in a boat and returned without him. Archimedes came very close to inventing calculus but couldn't make the final conceptual leap because the Ancient Greeks did not believe zero. The idea of using limits to get a result and getting an irrational number from an infinite sum of rational numbers would have been quite controversial.
This is exciting to hear. It's evident Professor Padilla is passionate about these breakthroughs. Keep up the good work, Brady. Pete, your animations have been a game changer for this channel.
That's not true, it always have been a mixture of both hard and easy topics. Take the last 6 videos, for example, I would argue 3 are very "simple"/"easy" ("Making a klein bottle", "a hairy problem" and "cow-culus")
okay but real talk this dude's been with numberphile since the beginning and HASN'T AGED A DAY Vampire? Fountain of Youth? Made a dark pact with the heathen maths Gods? Take your bets
I am myself mathematician but doing topics far from these mathematics, and I feel really impressed by the incredible pedagogical skill of this mathematician ! Thank you Tony !
From what I've heard it seems that unfortunately, the paper contains a mistake. It might be that Zhang or someone else will fix it, but it could be that it just can't be fixed. Also, at 8:22 Tony says that if you can find a Siegel zero then the twin prime conjecture will be proven. It's not quite as simple as finding a single Siegel zero. The definition of Siegel zeros has this constant c in it, and for Heath-Brown's theorem you need to prove that for all possible values of c>0, there exists a Siegel zero.
Zhang such an inspiration, he clearly devoted his life to humble steady hard work. I wonder if anyone who loves math and works hard can eventually contribute to the world even if they aren’t naturally talented
I don't know what the fancy character is used to depict a lower-case greek chi in the animation, but it definitely ain't a lower-case chi... EDIT: it seems to be the greek equivalent of "&", dubbed "kai" (same pronunciation as Pr. Padilla's chi). Still wrong character, but leading to an interesting discovery in ancient abbreviations!
probably a mistake by whoever typeset the animation. the hand written letter is chi and as far as i can see that's the standard notation as well. interesting to see this letter tho; it's new to me.
@@heavenlyactsatheavycost7629 ϗ is a ligature for the Greek word "και," which means "and"! It's similar to how the ampersand (&) is a ligature "et," the Latin word for "and."
Interesting stuff! One interesting corollary of the last point about Riemann Zeta tying into physics is that if a physics experiment behaves in an unexpected way in, it could be due to a failure of understanding the mathematics and not a failure of the theory itself. Or in other words, if there's a weird experimental result that relies on certain interpretation of underlying mathematics, that could develop the mathematical theory as well.
This needs a health warning! There are so many rabbit holes that are signposted in this video, all of which look as if they would be fun to follow up. A second health warning for being reminded that theories about primes link up to the sum of an infinite series of complex powers of numbers. Dangerous stuff - keep it coming.
Isn't the non existence fo Siegel zeros "the same" thing as the Riemann hypothesis? It just feels like that the critical line (going from 0 to infinity) now "just" is projected onto that line part going from c/log(D) to 1.
I think it would only disprove the generalised one, since we would know there's some generalised zeta function that has a non-trivial zero off the line, but it doesn't show that there's a non-trivial zero off the line on the original zeta function
In addition, the Riemann zeta function doesn't have real zeros inside the critical strip, so all of its non-trivial zeros are complex (i.e. not purely real). See e.g. Titchmarsh book on the RZF, p.30. Chapter 2, Section 12. Although the RZF can't have Siegel zeros, this doesn't imply a thing about the original RH either, for there still could be off the line complex zeros somewhere inside the critical strip.
I love the way he says, at 6:00, "we don't want to go into all the details here..." when in fact he completely lost me about 4 minutes ago. And the video still has 10 more minutes to run.
Except Heath-Brown's theorem will almost certainly will never prove the twin prime conjecture, because the Riemann Hypothesis is widely believed to be true.
Siegel primes would essentially guarantee very large fluctuations in the sequence of prime numbers, so much so that primes would inevitably need to be close together every so often. However, fluctuations of the primes appear to be FAR smaller than even the Riemann Hypothesis guarantees, so this method will almost certainly not prove the Twin Prime conjecture.
But if I understand correctly, Heath-Brown's theorem states that if there are no Siegel Primes then the Twin Prime Conjecture is false. And they said it's widely believed that these zeroes don't exist. So doesn't that mean that it's also believed the Twin Prime Conjecture is false?
@@VoodoosMaster I think the inexistence of Siegel Zeros doesn't prevent the Twin Prime conjecture to be true. 08:05 The statement says that one of them has to be true, meaning that at least one of them is true if the other is false (but maybe both are true, hahaha). However, both being false is not possible according to the theorem.
ϗ is the ligature for the Greek word "καί" which means "and." It is similar to the ampersand "&" in English, which is a ligature for "et," the Latin word for "and."
Any YT links for the relationship between the Riemann zeros and "energy levels of heavy nuclei" that Tony talked about? My searches are not getting anywhere. TIA!🙏🙏
8:26 isn't it the existence of infinite siegel zeros (one for each dirichlet character) that implies the twin prime conjecture which Roger Heath-Browns theorem says?
Yes - and technically speaking, the concept of "a Siegel zero" is not well-defined (you can always choose a small enough constant c so that any given zero is more than c/logD away from 1). You need an infinite collection of zeros that converge to 1 very rapidly in order to call the whole set a *collection* of siegel zeros.
The formulation at 8:12 is a bit unfortunate, as it can be read as if only one of the two statements is true, which is not what the Heath-Brown theorem states. It states that at least one of those statements is true.
To reiterate my questions about the -1/12… why is shifting a duplicated series underneath by one allowed or taken versus any other equal foul? Since when do we take an average of answers when a function gives more than one?!?!
Zhang did essentially the same thing as before. With the twin-prime conjecture he proved that there are infinitely many pairs of primes that differ by a number greater than 2 (so not exactly 2), and here he proved that there is a region where there are no Siegel Zeros, but that is smaller than needed for the full proof. I think this is the death knell for the existence of Siegel Zeros (if the proof holds of course).
@@oldvlognewtricks Thanks for that! That will teach me to be more careful when using expressions! Well, English is my second language... I've corrected the error because it distracted from the point I try to make...
I did n-t quite understand that if it'd disprove the „Generalized Riemann Hypothesis“, it'd disprove the „Riemann Hypothesis“ as well - as I did not understand if the RH is „totally“ included in the GRH or if it is just one case of the GRH and those „Siegel-zeros“ could be found to be in other cases but not in the „special“ case of the RH. Could some-one help? I'd appreciate it 🌞👍🏻
How come those simply hypothesis get some rather large and unexpect upper bounds? Or is that just a proof that all numbers are equal and we are bias towards smaller number?
1:45 A little nitpicking but that's a xi Tony writes there, not a zeta xD And at 5:22 the graphic uses a kappa instead of a chi, which Tony says and writes.
The Heath-Brown theorem claims that "ATLEAST one of the two is true". So if we do manage to find a Siegel zero, that would imply Twin prime conjecture is true. But, if we prove that Siegel zero does not exist (which it most probably does not), won't prove or disprove Twin prime conjecture.
I am fairly convinced that it should be like this about Siegel zeros: a) For each real Dirichlet character the corresponding L-function has at most 1 Siegel zero. b) Heath-Brown proved if there are infinite Siegel zeros (meaning for each real Dirichlet character one), then the twin prime conjecture is true. So the existence of one Siegel zero does not prove the twin prime conjecture.
It's impossible to have one Siegel zero- you just lower the constant until it isn't a Siegel zero. You need the infinite family to eliminate all possible constants.
@@btf_flotsam478 I guess this boils down on the definition of a Siegel zero. How do you define it? Is the Siegel zero defined in terms of multiple L-functions (meaning it's a zero for all L(s,\chi_q) for any \chi_q) or is it defined for one single Dirichlet L-function? I thought any exceptional zero that we can find for one specific Dirichlet L-function was called a Siegel zero and then we look at the collection of these zeros (for all Dirichlet characters) to formulate Heath-Browns theorem. Or do you call these just zeros and then define the Siegel zero to be one zero for all L(s,\chi_q)?
I may have missed something, but why is the existence of Siegel Zeros so hard to be proved or disproved? As I understood, those roots are real ones (no imaginary part), what makes them (in theory) somewhat easy to be located numerically (should they exist). Is the zeta function wildly oscillatory in the neighbourhood of 1 (approaching by the left)? Maybe I'm too crude on this topic.
I don’t know why, but I thought it was funny when he said the mathematician proved that there was an infinite amount of primes that differ by 70 million.
If Generalized Riemann Hypothesis is proven to be right, then "ordinary" Riemann Hypothesis would automaticaly be proven right too, if I'm correct. Sure. But GRH could be proven false while RH could still be true, right ? So, for example, there could be a Siegel zero AND Riemann Hypothesis could nonetheless be true (that would be an interesting possibility, IMHO). Else there would be no point in distinguishing between the two conjectures, for what I understand.
So if i understood this correctly, the existence of Siegel-zeroes doesn't disprove the Riemann-hypothesis, but the generalized Riemann-hypothesis. So the Riemann-hypothesis could still be true.
As far as I understood, the Riemann-hypothesis is a special case of the generalized one. If you disprove the generalized one, you disprove every one of its special cases, so the Riemann-hypothesis is then false.
Please someone help me to understand. The existance of siegel zero means the truth of twin prime conj, and RH is false, so is the truth of twin prime conj equivalent to the false RH?
If a Siegel zeroes is found (in general), it doesn't mean the RH is false, necessarily. It would imply the Generalized Riemann Hypothesis is false, but the (specific) Riemann Hypothesis may still be true even if the generalized version is false. Now, if a Siegel zero were found for the actual specific Riemann zeta function, it would prove RH false, and that would immediately imply that the twin prime conjecture is true. But no, RH and TPC are not necessarily equivalent. The theorem mentioned in the video states that the nonexistence of Siegel zeros and the twin prime conjecture cannot _both_ be false, but it is conceivable that both are _true._ If you were able to show one of the two things were false, then that would automatically imply the other is true, since both cannot be false. However, showing one of them to be true would tell us nothing about the other.
I am not sure I even want to ask how exactly you get to a proof that says "either there are no segel zeros or the twin prime conjecture must be true". Did not know some theorems where playing by Highlander rules.
Thank you very much for this video. Most of the articles I read about this were written very poorly and were hard to actually figure out what was going on.
Does it mean that if the Riemann hypothesis is true then the twin prime conjecture is false? (because if the Riemann hypothesis is true then there's no Siegel Zeros)
@@Peregringlk I can understand why one might think it is an exclusive or, but it is not an exclusive or. The theorem uses an inclusive or. If you read Terence Tao's blog post about it, he states the theorem as saying "at least one of the following is true".
"Trivial" in this phrase probably means something like "a PhD student in number theory can prove this as an exercise", not a common usage of that word.
It's fairly simple. Instead of the Riemann zeta function Sum (1/n^s) we look at the functions Sum (f(n)/n^s) for some additional function f(n) called Dirichlet character. If one chooses f(n):=1 then we get the Riemann zeta function.
Keep all the works of Hercules but replace him with an elder dude with negligible physical abilities. That's what Prof. Zhang is. And still he did it, and not just once. Amazing dude.
Instead of D it should be log(q) where q is the modulus of the Dirichlet character and c is just a constant. Both can be derived as upper bounds. If I remember correctly the theorem is due to Gronwall and Titchmarsh. It's a theorem about real and complex Dirichlet characters for L-functions.
I can’t get this off my mind. There is this problem I’ve been thinking about that involves a 4 digit combination lock. Like everyone who has ever used this type of lock they scramble all the numbers after locking it back up. What if every night, I could only change 1 (or 2, or 3) of the numbers on the lock. Is there a way to figure out my combination if a thief went to my lock after every scramble and collected data of what numbers were on my lock. Could he ever be 100% certain after a certain over a long period of time? Not sure how to even attempt solving this.
Well, if you can only change one number the thief knows 3 of the numbers, he just doesn't know which one, the next scramble if you changed another number he will know all 4 of them, if you always change the same number we can assume you do it randomly, and the correct one will never appear, so he knows what doesn't appear, keep calculating
I'm guessing it depends on whether or not you deliberately avoid at least one other combination to the one that opens the lock. If you don't, given enough time, you'll have cycled through every combination other than the one needed
Just a comment and someone please correct me if I'm wrong but I think you've misstated what Zhang proved. I thought he'd proved that there are an infinite number of primes that differ by 70 million OR LESS. Similarly I thought the proof regarding gaps of 246 is that there are an infinite number of primes that differ by 246 OR LESS.
8:20 : "if you find a Siegel zero, the twin prime conjecture has to be true". 11:45 : "i like to be one (Siegel zero), because then it disproves the Riemann hypothesis". These two statements contradict each other.
It should correctly be: - "if you find infinite Siegel zeros (one for each Dirichlet character chi mod q), then the twin prim conjceture is true." - "if you find a Siegel zero for the principal Dirichlet character, then you have found a zero for the Riemann zeta function above the 1/2 line (since there is an identity) and thus disproved the RH"
Does "an infinite number of primes that differ by 246 or less" mean the same as, every single prime number is only at most 246 numbers away from the next prime number?
AWESOME EXPLANATION !!! when this result first came out all the "science news" articles about it had no decent explanations, they were all for the mathematically illiterate and basically useless. MANY THANKS for doing this one !!! Also, I'm familiar with the fact that there are a great many pseudo-theorems of the sort "If the RH (or GRH) is true then XYZ", but was not aware there are any "if the RH (or GRH) is false then XYZ", describing some of the theorems in these two possible alternate worlds would be another great topic for you to cover.
See brilliant.org/numberphile for Brilliant and get 20% off their premium service (episode sponsor)
Order Tony's book Fantastic Numbers and Where to Find Them: A Cosmic Quest from Zero to Infinity
Amazon US - amzn.to/3JYQbws - Amazon UK - amzn.to/3M3yvB8
😅😅😅😅😊😅00
Ooo😊😊😊poooo
6:55 😅😅😅😊😊😊😊😊 7:01 7:02
Po99oooo😅ooo😊99889ppp😊p😅oo99😊😊😅9😊😊9😊o😊9😊 12:03 oo😊o😊ooo😊oooo9ooo0😊ook o9😅op9ook 😊o99😊9p99popolice p😊😊9😅
The nature of humanity is just that every so often someone accidentally invents the Riemann Hypothesis again.
😂😂
🦀
yep lol
It's sort of like how π keeps showing up even when you don't see a circle anywhere near.
@@namelastname4077 You can spend all your time contemplating the miseries of life and inevitablility of death if you want - personally I prefer to spend mine getting excited about fun cool things
I love Tony's tongue-in-cheek statement "without any controversy at all, it is equal to -1/12" 🤣
Once again making people think its normal summation, but its not
Let's be clear about this. 1+2+3+... does not equal -1/12. The series is the result of a function definition that doesn't work at -1. However, it's true that there is another more complicated function definition that gives the same values where the first definition works, and also works at -1. It's that other function that has the value -1/12 at -1.
A theoretical physicist tries to calculate something and gets the result 1+2+3+... . They guess that maybe they used the wrong maths, and maybe the right maths would give that other function so the answer is -1/12. If experiments then agree with this prediction the physicist becomes famous; if not they shrug and try a different way to calculate it.
Edited : I typed +1 when I meant -1. Hey ho.
@@john_g_harris What 1+2+3+... equals, depends on your particular choice of how to assign values to infinite series. It's not possible to assign any finite value to it if you choose to adopt the standard definition but there are other definitions. The Ramanujan summation of 1+2+3+... does equal -1/12. Which particular definition is relevant to solving any particular problem can vary depending on the context in which the summation arises.
Classic..
@@john_g_harris Regularization of the Riemann zeta function at s = -3 is used in calculating the Casimir effect and more generally in quantum mechanics there's a fair amount of renormalization where techniques are used to get a finite sum from a divergent series to get actual results.
The argument that the sum of 1+2+3+ ... does not equal -1/12 because it uses a different method of getting the result comes up a lot. While it's important to recognize that yes, it doesn't mean "equals" in the same way as other "equals", this exact sort of thing has happened before. By the rules of basic arithmetic, the sum of a rational number and another rational number is a rational number. But take all the nonnegative integers and sum the reciprocal of their factorials and you get the transcendental number e. However, getting to this result, or for that matter getting the result of any convergent infinite series requires a different technique than basic arithmetic.
This is not a controversial result today because people are used to the concept of limits and zero, but in the time of Pythagoras or Archimedes, it would have been jus as controversial as summing the positive integers to -1/12. There's an apocryphal story that a member of the Cult of Pythagoras came up with a proof that the square root of 2 is irrational and that the Pythagoreans were so incensed with the result because it broke the rules that they believed in that they took him out to sea in a boat and returned without him. Archimedes came very close to inventing calculus but couldn't make the final conceptual leap because the Ancient Greeks did not believe zero. The idea of using limits to get a result and getting an irrational number from an infinite sum of rational numbers would have been quite controversial.
This is exciting to hear. It's evident Professor Padilla is passionate about these breakthroughs. Keep up the good work, Brady. Pete, your animations have been a game changer for this channel.
I've never been more confused by land-owls and seagulls, but I'm glad he's excited about them.
Yitang Zhang is like a more successful version of Matt Parker. He makes breakthroughs in important cases, but not to the point that was conjectured.
So you’re saying Matt is kind of a Parker Yitang Zhang
@@TimMaddux exactly
🤣🤣🤣
He *makes *breakthroughs
@@ophello Haha, thanks.
Man, Numberphile has covered all of the simple math topics. These kinds of videos are HEAVY
I finally feel a little bit better seeing someone else feel the same.
That's not true, it always have been a mixture of both hard and easy topics. Take the last 6 videos, for example, I would argue 3 are very "simple"/"easy" ("Making a klein bottle", "a hairy problem" and "cow-culus")
I recommend the 3Blue1Brown video on the riemann zeta hypothesis for background here. It is visually beautiful.
“This is HEAVY, doc” -Marty McFly
Wow it’s Verlisify! The search for Siegel zeroes so hard they call it Verlisify. Verlisify isify whoo whoo
In the future we'll refer to "Zhang Numbers" : arbitrary values that allowed us to make headway in various proofs.
okay but real talk this dude's been with numberphile since the beginning and HASN'T AGED A DAY
Vampire? Fountain of Youth? Made a dark pact with the heathen maths Gods? Take your bets
His age is a mathematical constant, rather than a variable.
idk man, hes aged a bit since his smosh days
He's asymptotically aging
He is probably a Youkai lol
A healthy even diet, with an odd snack here and there
I am myself mathematician but doing topics far from these mathematics, and I feel really impressed by the incredible pedagogical skill of this mathematician ! Thank you Tony !
Very astute product placement, Tony! I ordered your book when it was originally announced on Numberphile and thoroughly enjoyed it.
From what I've heard it seems that unfortunately, the paper contains a mistake. It might be that Zhang or someone else will fix it, but it could be that it just can't be fixed.
Also, at 8:22 Tony says that if you can find a Siegel zero then the twin prime conjecture will be proven. It's not quite as simple as finding a single Siegel zero. The definition of Siegel zeros has this constant c in it, and for Heath-Brown's theorem you need to prove that for all possible values of c>0, there exists a Siegel zero.
What's the source on that first bit? How critical is the mistake?
Remember that Wiles’ proof of Fermat’s Last Theorem had a mistake. Give it time and we’ll see.
Can we all appreciate how the style of video hasn't changed in forever.
Yitang is an absolute genius and a legend
A link between the Twin Prime Conjecture and the Reimann Hypothesis? Numberphile really knows how to stop me working on my thesis!
Same here! My thesis is in fluid dynamics but this is way more interesting to me
There are already connections. By the nature of the riemann zeroes generating the prime number theorem, you get twin prime conjecture somewhat easily.
What's your thesis on? Hope you are finding it interesting.
Zhang such an inspiration, he clearly devoted his life to humble steady hard work. I wonder if anyone who loves math and works hard can eventually contribute to the world even if they aren’t naturally talented
What r u talking about?
Yes you can !
Do it if you love math
If love it it's possible, if you still have doubt then watch David goggins Then if you still don't go after it you will regret it
@@gauravbharwan6377 You wanna be a mathematician too, bro?
R u kidding me? This is number theory. Ofc he's very talented. He was concidered the best back in the school
I like this channel alot, its better than white noise and helps me sleep. No joke, super helpful.
I don't know what the fancy character is used to depict a lower-case greek chi in the animation, but it definitely ain't a lower-case chi...
EDIT: it seems to be the greek equivalent of "&", dubbed "kai" (same pronunciation as Pr. Padilla's chi). Still wrong character, but leading to an interesting discovery in ancient abbreviations!
It's a mistake. ϗ is the ligature for the Greek word "kai" which means "and." It is similar to the ampersand "&" in English.
probably a mistake by whoever typeset the animation. the hand written letter is chi and as far as i can see that's the standard notation as well. interesting to see this letter tho; it's new to me.
@@heavenlyactsatheavycost7629 ϗ is a ligature for the Greek word "και," which means "and"! It's similar to how the ampersand (&) is a ligature "et," the Latin word for "and."
Padilla also wrote a script xi when he should have written zeta.
@@jaoswald Probably thinking of the completed zeta function.
Imagine mathematicians were like song artists.
Twitter post: "New RH proof dropping on December 21st, 7 PM EST. Don't miss it"
This actually does happen on sites like Math Overflow
Hey, dude, check this out! This stuff is fire! Read it while on shrooms, it will blow your mind!
There is a couple schizoids who keep dropping a "proof" every couple months
Brilliant, I love videos like this about ongoing math developments
More videos like this, please! This was fantastic.
Interesting stuff! One interesting corollary of the last point about Riemann Zeta tying into physics is that if a physics experiment behaves in an unexpected way in, it could be due to a failure of understanding the mathematics and not a failure of the theory itself.
Or in other words, if there's a weird experimental result that relies on certain interpretation of underlying mathematics, that could develop the mathematical theory as well.
This needs a health warning!
There are so many rabbit holes that are signposted in this video, all of which look as if they would be fun to follow up.
A second health warning for being reminded that theories about primes link up to the sum of an infinite series of complex powers of numbers.
Dangerous stuff - keep it coming.
It truly is fascinating how long number theory reaches into other fields of mathematics in order to even begin to grasp the nature of primes
He's so proud of the video from 10 years ago, he still has the 2012 calendar.
Isn't the non existence fo Siegel zeros "the same" thing as the Riemann hypothesis? It just feels like that the critical line (going from 0 to infinity) now "just" is projected onto that line part going from c/log(D) to 1.
No one’s gonna talk about the fact his mouse is plugged into the wall socket?
nice one but of course it's plugged into the keyboard
One thing is for sure. Yitang Zhang is a beast!
at 1:46, Tony writes what looks like a Xi instead of a Zeta. Am I wrong?
If a "Siegel Zero" is found or proven to exist, is it "only" the "Generalized Riemann hypothesis" that fails or also the normal "Riemann hypothesis" ?
I think it would only disprove the generalised one, since we would know there's some generalised zeta function that has a non-trivial zero off the line, but it doesn't show that there's a non-trivial zero off the line on the original zeta function
If it is a Siegel zero for one Dirichlet character it doesn't mean automatically it is one for another.
In addition, the Riemann zeta function doesn't have real zeros inside the critical strip, so all of its non-trivial zeros are complex (i.e. not purely real). See e.g. Titchmarsh book on the RZF, p.30. Chapter 2, Section 12.
Although the RZF can't have Siegel zeros, this doesn't imply a thing about the original RH either, for there still could be off the line complex zeros somewhere inside the critical strip.
12:07
"Seagul" Zero is in quantum state.
Now we have "Seagul" Zero and Schrodingers Cat.
I love the way he says, at 6:00, "we don't want to go into all the details here..." when in fact he completely lost me about 4 minutes ago. And the video still has 10 more minutes to run.
Wow. I mainly love how this can prove the twin prime conjecture to be true. Its very exciting actually
Except Heath-Brown's theorem will almost certainly will never prove the twin prime conjecture, because the Riemann Hypothesis is widely believed to be true.
Siegel primes would essentially guarantee very large fluctuations in the sequence of prime numbers, so much so that primes would inevitably need to be close together every so often. However, fluctuations of the primes appear to be FAR smaller than even the Riemann Hypothesis guarantees, so this method will almost certainly not prove the Twin Prime conjecture.
But if I understand correctly, Heath-Brown's theorem states that if there are no Siegel Primes then the Twin Prime Conjecture is false. And they said it's widely believed that these zeroes don't exist. So doesn't that mean that it's also believed the Twin Prime Conjecture is false?
@@VoodoosMaster I think the inexistence of Siegel Zeros doesn't prevent the Twin Prime conjecture to be true. 08:05 The statement says that one of them has to be true, meaning that at least one of them is true if the other is false (but maybe both are true, hahaha).
However, both being false is not possible according to the theorem.
@@jagatiello6900 Ohhh got it, thank you. Then it's not as exciting as I imagined lol
When D=1, ie the twin prime conjecture, c/Log(1) is undefined. Where would I check to find a Seigel zero?
5:20 ive never seen a chi written like that before
ϗ is the ligature for the Greek word "καί" which means "and." It is similar to the ampersand "&" in English, which is a ligature for "et," the Latin word for "and."
Thank you, your videos are always well worth the time to watch!
Any YT links for the relationship between the Riemann zeros and "energy levels of heavy nuclei" that Tony talked about? My searches are not getting anywhere. TIA!🙏🙏
8:26 isn't it the existence of infinite siegel zeros (one for each dirichlet character) that implies the twin prime conjecture which Roger Heath-Browns theorem says?
Yes - and technically speaking, the concept of "a Siegel zero" is not well-defined (you can always choose a small enough constant c so that any given zero is more than c/logD away from 1). You need an infinite collection of zeros that converge to 1 very rapidly in order to call the whole set a *collection* of siegel zeros.
He managed to get through a whole 3 mins before mentioning Euler XD
What's with the shape of χ on the transcription around 5:17? I thought that it was a kappa before I saw the handwritten version.
If c is any number, isn't (c / log D) unbounded?
The formulation at 8:12 is a bit unfortunate, as it can be read as if only one of the two statements is true, which is not what the Heath-Brown theorem states. It states that at least one of those statements is true.
Finally something about zeta/l -functions
To reiterate my questions about the -1/12… why is shifting a duplicated series underneath by one allowed or taken versus any other equal foul? Since when do we take an average of answers when a function gives more than one?!?!
Zhang did essentially the same thing as before. With the twin-prime conjecture he proved that there are infinitely many pairs of primes that differ by a number greater than 2 (so not exactly 2), and here he proved that there is a region where there are no Siegel Zeros, but that is smaller than needed for the full proof. I think this is the death knell for the existence of Siegel Zeros (if the proof holds of course).
death knell*
“death nail” made me laugh… Mutant offspring of “death knell” and “nail in the coffin” 😂
@@oldvlognewtricks Thanks for that! That will teach me to be more careful when using expressions! Well, English is my second language... I've corrected the error because it distracted from the point I try to make...
I did n-t quite understand that if it'd disprove the „Generalized Riemann Hypothesis“, it'd disprove the „Riemann Hypothesis“ as well - as I did not understand if the RH is „totally“ included in the GRH or if it is just one case of the GRH and those „Siegel-zeros“ could be found to be in other cases but not in the „special“ case of the RH.
Could some-one help? I'd appreciate it 🌞👍🏻
I ve been waiting a month for this!
How come those simply hypothesis get some rather large and unexpect upper bounds?
Or is that just a proof that all numbers are equal and we are bias towards smaller number?
I'd have to watch this video Graham's number of times to fully understand it
1:45 A little nitpicking but that's a xi Tony writes there, not a zeta xD And at 5:22 the graphic uses a kappa instead of a chi, which Tony says and writes.
2024 in the answer makes me think this is an Olympiad question 2 years from now.
😂😂😂😂
Why cant this be checked by a computed? Is the function wildly oscillatory there or something?
The Heath-Brown theorem claims that "ATLEAST one of the two is true".
So if we do manage to find a Siegel zero, that would imply Twin prime conjecture is true.
But, if we prove that Siegel zero does not exist (which it most probably does not), won't prove or disprove Twin prime conjecture.
It came to me the thought that the Riemann-hypothesis could become the equivalent of the fifth Euclidean postulate but for number theory.
But if you look at the graph of zeta on the real graph there is no positive point where where zeta is zero
Or those zeros doesn't exist when D is 1 ?
I am fairly convinced that it should be like this about Siegel zeros:
a) For each real Dirichlet character the corresponding L-function has at most 1 Siegel zero.
b) Heath-Brown proved if there are infinite Siegel zeros (meaning for each real Dirichlet character one), then the twin prime conjecture is true.
So the existence of one Siegel zero does not prove the twin prime conjecture.
It's impossible to have one Siegel zero- you just lower the constant until it isn't a Siegel zero. You need the infinite family to eliminate all possible constants.
@@btf_flotsam478 I guess this boils down on the definition of a Siegel zero. How do you define it? Is the Siegel zero defined in terms of multiple L-functions (meaning it's a zero for all L(s,\chi_q) for any \chi_q) or is it defined for one single Dirichlet L-function? I thought any exceptional zero that we can find for one specific Dirichlet L-function was called a Siegel zero and then we look at the collection of these zeros (for all Dirichlet characters) to formulate Heath-Browns theorem.
Or do you call these just zeros and then define the Siegel zero to be one zero for all L(s,\chi_q)?
I may have missed something, but why is the existence of Siegel Zeros so hard to be proved or disproved? As I understood, those roots are real ones (no imaginary part), what makes them (in theory) somewhat easy to be located numerically (should they exist). Is the zeta function wildly oscillatory in the neighbourhood of 1 (approaching by the left)? Maybe I'm too crude on this topic.
This is something I've got to watch again. But not tonight.
I don’t know why, but I thought it was funny when he said the mathematician proved that there was an infinite amount of primes that differ by 70 million.
Oh boy! Any advancement in number theory involving Riemann excites me.
Did you know the following fact about Riemann and primes: Riemann's hands each had a prime number of fingers!
@@u.v.s.5583 Using the term "digits" would have been more correct and a double entendre. Missed opportunity.
Okay, so what is bigger, the last two twin primes or tree(3)?
If Generalized Riemann Hypothesis is proven to be right, then "ordinary" Riemann Hypothesis would automaticaly be proven right too, if I'm correct.
Sure. But GRH could be proven false while RH could still be true, right ? So, for example, there could be a Siegel zero AND Riemann Hypothesis could nonetheless be true (that would be an interesting possibility, IMHO).
Else there would be no point in distinguishing between the two conjectures, for what I understand.
That is correct
If the GRH is true, the RH is true
But the converse is not the case.
So if i understood this correctly, the existence of Siegel-zeroes doesn't disprove the Riemann-hypothesis, but the generalized Riemann-hypothesis. So the Riemann-hypothesis could still be true.
And, of course, his work supports the generalised Riemann Hypothesis anyway.
No wrong
As far as I understood, the Riemann-hypothesis is a special case of the generalized one. If you disprove the generalized one, you disprove every one of its special cases, so the Riemann-hypothesis is then false.
Prof Tony luvs his numbers
The sum of all integers is -1/12? In which universe?
What's even more interesting to me is how did they find a link between twin prime and This function? That's what I want to know about
I can only understand like 10% of the whole video. Still watch it
If c us arbitrary then can't the width be any size?
I like that this guy is embracing having the most controversial numberphile video
Knowing that, how many pots of paint does Paul need to paint his wall?
How many watermelons did Matt have?
Steven Siegel is an amazing world class action zero.
1:47 I believe that's the letter xi, not zeta.
Zeta is the name of the function, not the character
@@dancurtis8476 the function got its name from the character...
Please someone help me to understand. The existance of siegel zero means the truth of twin prime conj, and RH is false, so is the truth of twin prime conj equivalent to the false RH?
If a Siegel zeroes is found (in general), it doesn't mean the RH is false, necessarily. It would imply the Generalized Riemann Hypothesis is false, but the (specific) Riemann Hypothesis may still be true even if the generalized version is false. Now, if a Siegel zero were found for the actual specific Riemann zeta function, it would prove RH false, and that would immediately imply that the twin prime conjecture is true.
But no, RH and TPC are not necessarily equivalent. The theorem mentioned in the video states that the nonexistence of Siegel zeros and the twin prime conjecture cannot _both_ be false, but it is conceivable that both are _true._ If you were able to show one of the two things were false, then that would automatically imply the other is true, since both cannot be false. However, showing one of them to be true would tell us nothing about the other.
1:45 what does he mean by that?
I am not sure I even want to ask how exactly you get to a proof that says "either there are no segel zeros or the twin prime conjecture must be true". Did not know some theorems where playing by Highlander rules.
In the heath Brown conjecture, can both be true?
Thank you very much for this video. Most of the articles I read about this were written very poorly and were hard to actually figure out what was going on.
Does it mean that if the Riemann hypothesis is true then the twin prime conjecture is false? (because if the Riemann hypothesis is true then there's no Siegel Zeros)
No because OR can mean one or both are true.
@@pmcate2 He seems to imply an exclusive or.
@@Peregringlk Where?
@@Peregringlk I can understand why one might think it is an exclusive or, but it is not an exclusive or. The theorem uses an inclusive or. If you read Terence Tao's blog post about it, he states the theorem as saying "at least one of the following is true".
Can someone explain how the negative even number values for the real part of s yield zeros and why this is trivial?
"Trivial" in this phrase probably means something like "a PhD student in number theory can prove this as an exercise", not a common usage of that word.
"There's a more general version of the Riemann hypothesis called the generalized Riemann hypothesis. It's the Riemann hypothesis but generalized."
It's fairly simple. Instead of the Riemann zeta function Sum (1/n^s) we look at the functions Sum (f(n)/n^s) for some additional function f(n) called Dirichlet character. If one chooses f(n):=1 then we get the Riemann zeta function.
No, it is so called in honor of the famous mathematician Bernhard Generalized Riemann (1967-1975)
Anything new about whether or Not Zhang is rigth?
I was confused for a moment because I was conflating the Twin Prime Conjecture with the Riemann Conjecture.
0:45 Why does he have a 10 year old calendar hanging on the wall? And who is Jessica?
Yitang Zhang is the modern day Hercules of mathematics.
Keep all the works of Hercules but replace him with an elder dude with negligible physical abilities. That's what Prof. Zhang is. And still he did it, and not just once. Amazing dude.
this is so difficult to understand.. where does the D and the c come from? What does the generalized function mean?
Instead of D it should be log(q) where q is the modulus of the Dirichlet character and c is just a constant. Both can be derived as upper bounds. If I remember correctly the theorem is due to Gronwall and Titchmarsh. It's a theorem about real and complex Dirichlet characters for L-functions.
I can’t get this off my mind. There is this problem I’ve been thinking about that involves a 4 digit combination lock. Like everyone who has ever used this type of lock they scramble all the numbers after locking it back up. What if every night, I could only change 1 (or 2, or 3) of the numbers on the lock. Is there a way to figure out my combination if a thief went to my lock after every scramble and collected data of what numbers were on my lock. Could he ever be 100% certain after a certain over a long period of time? Not sure how to even attempt solving this.
Well, if you can only change one number the thief knows 3 of the numbers, he just doesn't know which one, the next scramble if you changed another number he will know all 4 of them, if you always change the same number we can assume you do it randomly, and the correct one will never appear, so he knows what doesn't appear, keep calculating
I'm guessing it depends on whether or not you deliberately avoid at least one other combination to the one that opens the lock. If you don't, given enough time, you'll have cycled through every combination other than the one needed
Just a comment and someone please correct me if I'm wrong but I think you've misstated what Zhang proved. I thought he'd proved that there are an infinite number of primes that differ by 70 million OR LESS. Similarly I thought the proof regarding gaps of 246 is that there are an infinite number of primes that differ by 246 OR LESS.
Love the office window
Mochizuki may have proved non-existence of Siegel zeros.
Does it mean that the twin prime conjuncture is false? I don't think the Riemann's hypothesis is false.
There is an elementary statement about the RH related to the growth of the mertens function.
And why do you keep looking upwards???
What does zero divided by zero equal?
“The jury is still out!” 😊
8:20 : "if you find a Siegel zero, the twin prime conjecture has to be true". 11:45 : "i like to be one (Siegel zero), because then it disproves the Riemann hypothesis". These two statements contradict each other.
It should correctly be:
- "if you find infinite Siegel zeros (one for each Dirichlet character chi mod q), then the twin prim conjceture is true."
- "if you find a Siegel zero for the principal Dirichlet character, then you have found a zero for the Riemann zeta function above the 1/2 line (since there is an identity) and thus disproved the RH"
I didn't understand half of that but I'm happy for the progress on Riemann hypothesis :)
Watching this in 2024
Does "an infinite number of primes that differ by 246 or less" mean the same as, every single prime number is only at most 246 numbers away from the next prime number?
I don't think so, there could be primes without 246 pairs as well
No. This says just that there are infinite pairs of primes (p,p+D) where D
AWESOME EXPLANATION !!! when this result first came out all the "science news" articles about it had no decent explanations, they were all for the mathematically illiterate and basically useless. MANY THANKS for doing this one !!!
Also, I'm familiar with the fact that there are a great many pseudo-theorems of the sort "If the RH (or GRH) is true then XYZ", but was not aware there are any "if the RH (or GRH) is false then XYZ", describing some of the theorems in these two possible alternate worlds would be another great topic for you to cover.
8:14 that’s not an exclusive or relationship for those wondering, both statements could be true, but at least one must be true.
8:15 so wait, does this mean that either the RH is true or the twin prime conjecture is true? As in either or they can't be both true?
No, probably both are true. What can't happen is that they're both false, that there are only finitely many twin primes and the GRH fails.
Chi 2 alternates between -1 and 1, not 0 and 1.