Google vs. WolframAlpha
Mar 2, 2010
I found this page in LiveJournal, claiming that Fermat’s Last Theorem is false. Wait a sec, Andrew Wiles spent 10 years trying to prove it, that can’t be right. So I did a little check using my dear old friend, WolframAlpha.
According to Google:
(7 706 288^20) + (7 937 911^20) – (8 114 575^20) = 0
But according to WolframAlpha:
(7706288^20) + (7937911^20) – (8114575^20)
= -527601188217067653783789971407748001551858809706200140816163…
…12814496451245102847590968160489012839789723084835696729248
And what do you know, here is a list of near-misses for Fermat’s Last Theorem.
Lesson learned: Don’t rely on Google too much. Haha.
3 Comments
leave one →
The perils of not using arbitrary precision big-integers.
Well, it shouldn’t be too much of a deal for Google because I don’t think many people use Google as a calculator when there’s Wolfram Alpha, as well as the regular calc.exe or whatever your system equivalent is.
Haha, sometimes I use Google for currency conversion. I’ve never tried using Google as a serious calculator, and I just learned NOT to use it. At least, for computations of the order 10^20.
That’s very Feynman-esque of you, Tinzky, to do that. Nice. I’m sure Richard Feynman is now chuckling at google… and stretching a ice-cold rubber o-ring.