Soon I will have the absolute correct formula for the prime counting function

>soon I will have the absolute correct formula for the prime counting function
>I will just this to prove Riemann zeta after cracking all the RSA I want and mining a shit ton of bit coins
Feels good man

Other urls found in this thread:

maa.org/sites/default/files/pdf/upload_library/22/Chauvenet/Zagier.pdf
twitter.com/NSFWRedditImage

If we can't have it, we don't care.

Well I'm not done yet. I have to finish it and prove it.

You can do more than that

Damn shed get rekt

>not stealing the said coins
wew

Fuck. Sauce? Would splerg on her tits mate.

bitcoin uses elliptic curve signatures based on discrete log not RSA

does integer factorization even help break the discrete log problem?

Kek, new year's no fap?

Update: I have failed for now. But this isn't the end.

>cracking all the RSA I want and mining a shit ton of bit coins

There are n/log(n) primes less than n. This is still exponential.

>CS majors once again failing to understand their own fucking subject

Sorry I'm halfway through EE. Obviously an explicit prime formula would cut down on sieving time, moron.

But wouldnt you piss on gemma artetons tits? Blimey.

Idk who that is but I probably would

It's still
Ω(n^0.99999999999999999999999
9999999999999999999999999999
9999999999999999999999999999
9999999999999999999999999999
9999999999999999999999999999
9999999999999999999999999999
9999999999999999999999999998)

hahaha good bait op sage

Really? You don't think the final equation (if we ever do find it) will still be as slow as making multiple comparisons? On the other hand, stuff like Wilson-Lagrange seem pretty simple but they aren't very fast.

the fuck are you saying?

there is 0 way you can prove anything with riemann in its name before figuring out p vs np

p vs np proof then riemann hypothesis proof then string theory then random sequence extrapolation proof

>there is 0 way you can prove anything with riemann in its name before figuring out p vs np
That's complete bullshit, in no way does RH depend on P vs NP. P vs NP will be the last of them to be proven. RH is a relatively straightforward hypothesis about the critical zeroes of an infinite product/series lying on a line. It will take a while for it to be proven, but we're much closer than P vs NP. There's a chance that a RH proof won't even include prime numbers.

Viva Espania

Just gimme sauce u cunt

Who is this?

>Obviously an explicit prime formula would cut down on sieving time, moron

Only by a factor of 1/2ln(n) which is worthless. Other factoring algorithms already have a much better speed up.

>Really? You don't think the final equation (if we ever do find it) will still be as slow as making multiple comparisons?

Even if the prime listing function is constant time, the prime number theorem says there are n/ln(n) primes less than n. So there are √n/ln(√n) primes less than √n to check as factors. Since the limit n^∊/ln(n) →∞ ∀∊>0; Θ(√n/ln(n)) is Ω(n^1/2-∊) for any arbitrarily small but nonzero ∊.

Fermat's factorization method is O(n^1/3) which has a much better speed up of a factor of 1/n^1/6 but still exponential in the bit length of n (aka log2(n)). And we have sub exponential (but super polynomial) methods likes the quadratic sieve which is O(e^c√[ln(n)ln(ln(n))]) and general number field sieve which is O(e^(c∛[ln(n)ln(ln(n))^2]))...

>random sequence extrapolation
We need to stop this meme.

>random sequence extrapolation proof
kek

gemma arterton. I'm sure she's great at math

Too bad Jake Gyllenhalovicz is already knee deep in that pussy.

Kehli Krumme

Doesn't it seem likely that a primes formula would lead to an increased understanding of factorization? All algorithms simply use guess and check right now disguised in various forms and with different optimizations. And the prime number theory of n/ln[n] is shaky at best, and it's been improved with li[x].
What I'm saying is that a fundamentally different type of algorithm might be on the horizon, with much more of a predictive angle rather than a guess-and-check. With a development like that I think it might shrink to something like O(1) even, but that's hard to say.

...

if you have a correct prime counting function thats great
it will tell you nothing about RH
and nothing about factorising

You aren't fully understanding what I'm saying. I don't have anything significant yet since my hopes were a dead end. But a true explicit prime counting function would likely also prove enough theory to explain factoring. And an explicit Pi[x] would probably be very helpful in RH

>There's a chance that a RH proof won't even include prime numbers.
Could you expand on this?

>tfw have intuitively solved it but can't write it down

sometimes I forget it and then solve it again lol

I'm not sure about anything but from what I've heard mathematicians say, it's very likely it will include primes, but maybe if it skipped over the Euler product for the Dirichlet series it's possible to skip them. Idk

Bs

>With a development like that I think it might shrink to something like O(1) even

It's going to have to read the input and take log2(n) time.

>And the prime number theory of n/ln[n] is shaky at best

It's proven.

maa.org/sites/default/files/pdf/upload_library/22/Chauvenet/Zagier.pdf

fucking gemma jesus

Are there any opinions among experts on what is the complexity of integer factorization?

just sell huge primes to the NSA and live life in luxury. no need to mine bitcoins

I'm pretty sure you are far from it, perhaps you just thought of the problem but have no clue what to do.
If you have something, tell us and let's collaborate and solve this.. try to at least

Primes have nothing to do with mining bitcoins unless you break SHA which is doubtful.

Yes I know I am far from it. I made my own series representation for the nth prime but that was a dead end. I was hopeful that the Wilson theorem would be helpful since n is prime iff (Gamma(n)+1)/n is an odd integer. I'd be down with collaborating but seeing as none of the math geniuses have solved this, I can't possibly make any advancements.

>Just sell huge primes to the NSA and then be tortured until you tell them how you got them and then killed to make sure nobody else can get them

oh shit

Whoa whoa, what's this? How much can I get per prime?

3.50

Hottest woman I've seen in my whole life.
10/10 body
Hips are perfect
Tits on point
Face speaks for itself
Thighs just thicc enough
Ass probably as nice and fuckable as it gets
pussy must feel like heaven

photoshop

here's another sauce for scientific analysis

She can't be that much worse in reality, she looks like her blowjobs would make me cum 3 times in 30 seconds and pussy so soft and tight it would pump out 3 years worth of semen in a night and let me see 4th Dimension.

I'm really horny btw

She is so hot. I'd love to fuck her while she holds my head to her bosom.

>not wanting cute bf to snuggle with
Gtfo user this is a gay website

Yes I thought the same thing when I first came across it. I realized it was a shopping and found both of them. Here's the original. Still pretty hot. Not sure about the fourth dimension though.