Timeline for What advantage humans have over computers in mathematics?
Current License: CC BY-SA 3.0
50 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 28, 2021 at 22:40 | answer | added | Fallen Apart | timeline score: 2 | |
| Aug 28, 2021 at 22:30 | review | Close votes | |||
| Sep 1, 2021 at 13:44 | |||||
| Aug 28, 2021 at 20:55 | answer | added | Anixx | timeline score: 0 | |
| Feb 26, 2019 at 13:35 | review | Close votes | |||
| Mar 7, 2019 at 3:05 | |||||
| Dec 11, 2017 at 0:29 | review | Close votes | |||
| Dec 11, 2017 at 15:11 | |||||
| Nov 5, 2017 at 11:30 | comment | added | yters | Could a computer derive the halting problem? | |
| Aug 3, 2017 at 20:43 | answer | added | Yochay Jerby | timeline score: 5 | |
| Aug 3, 2017 at 14:27 | answer | added | Alex Gavrilov | timeline score: 4 | |
| Aug 3, 2017 at 12:42 | review | Close votes | |||
| Aug 3, 2017 at 15:53 | |||||
| Apr 13, 2017 at 12:57 | history | edited | CommunityBot |
replaced http://mathoverflow.net/ with https://mathoverflow.net/
|
|
| Apr 15, 2016 at 7:03 | comment | added | Dominic van der Zypen | Don't panic about machines replacing mathematicians. This is a great NYtimes article on the multiple promises that AI has made in the past -- but hasn't been keeping the promises: mobile.nytimes.com/2016/04/07/science/… | |
| Apr 12, 2016 at 3:08 | review | Close votes | |||
| Apr 12, 2016 at 7:36 | |||||
| Mar 18, 2016 at 10:35 | comment | added | Philippe Gaucher | That machines (neural networks) could do mathematics one day does not mean that theorems produced by these machines will be interesting for the rest of the mathematical community. As a comparison, maybe a machine will be able to create a company in the future, that does not mean that it will find funding to build it. | |
| Mar 17, 2016 at 23:59 | comment | added | Kimball | Jordan Ellenberg briefly speculates about this in his book How not to be wrong. | |
| Mar 15, 2016 at 21:47 | answer | added | Douglas Zare | timeline score: 36 | |
| Mar 15, 2016 at 21:35 | comment | added | R. van Dobben de Bruyn | One day, computers will be better at imagining the future than humans. | |
| Mar 15, 2016 at 21:31 | answer | added | abo | timeline score: 6 | |
| Mar 15, 2016 at 19:20 | comment | added | Zach H | We can make mistakes (and that's only partially a joke). | |
| Mar 15, 2016 at 17:28 | answer | added | მამუკა ჯიბლაძე | timeline score: 7 | |
| Mar 15, 2016 at 16:38 | comment | added | Morteza Azad | There is a much older related question with a similar title on Quora. Some of the answers over there could be of your interest: What are the advantages that humans have over machines in trading financial markets? | |
| Mar 15, 2016 at 15:45 | history | reopened |
joro Gil Kalai Alex Degtyarev Andrey Rekalo Moritz Firsching |
||
| Mar 15, 2016 at 13:17 | review | Reopen votes | |||
| Mar 15, 2016 at 15:45 | |||||
| Mar 15, 2016 at 7:37 | history | closed |
Suvrit Marco Golla Stefan Kohl♦ András Bátkai Stefan Waldmann |
Not suitable for this site | |
| S Mar 15, 2016 at 5:17 | history | mod moved comments to chat | |||
| S Mar 15, 2016 at 5:17 | comment | added | Todd Trimble | Comments are not for extended discussion; in particular, the part of this conversation that is properly a meta concern has been moved to chat. | |
| Mar 14, 2016 at 20:12 | review | Close votes | |||
| Mar 15, 2016 at 7:37 | |||||
| Mar 14, 2016 at 20:05 | comment | added | reuns | the axiom of artificial intelligence is that the human brain is nothing more than a very efficient (well programmed, by the DNA..) algorithm. under this axiom, a computer could theoretically simulate a human brain, hence, human have no advantage over computers. in the same way, because humans can build machines to help them solving a task, computers (machines) have no advantage over humans. | |
| Mar 14, 2016 at 14:53 | history | protected | Douglas Zare | ||
| Mar 14, 2016 at 14:53 | history | reopened |
Mikhail Katz Wolfgang Fedor Petrov Brendan McKay Douglas Zare |
||
| Mar 13, 2016 at 19:14 | comment | added | J.-E. Pin | Independently of the pertinence of this question for MO, there seems to be a confusion between finding a proof and verifying a proof in this discussion. Proof verification is much more advanced (see Gonthier's recent proof of Feit–Thompson theorem ) than discovery of proofs by computers. | |
| Mar 13, 2016 at 17:34 | comment | added | user9072 | In case somebody want to discuss about the on-topicness of the question, there is a meta thread | |
| Mar 13, 2016 at 16:48 | history | edited | Mikhail Katz |
edited tags
|
|
| Mar 13, 2016 at 14:58 | review | Reopen votes | |||
| Mar 13, 2016 at 18:58 | |||||
| Mar 13, 2016 at 14:16 | history | closed |
Will Jagy Chris Godsil Franz Lemmermeyer Gerald Edgar Johannes Hahn |
Not suitable for this site | |
| Mar 13, 2016 at 13:29 | answer | added | Alexandre Eremenko | timeline score: 12 | |
| Mar 13, 2016 at 12:44 | answer | added | logicute | timeline score: 10 | |
| Mar 13, 2016 at 11:33 | history | made wiki | Post Made Community Wiki by Todd Trimble | ||
| Mar 13, 2016 at 11:30 | answer | added | Tony Huynh | timeline score: 12 | |
| Mar 13, 2016 at 10:09 | answer | added | Konstantinos Gaitanas | timeline score: 21 | |
| Mar 13, 2016 at 9:38 | comment | added | post.as.a.guest | Human mathematicians have a strong incentive not to develop automatic proof methodologies, as it would likely obviate the research aspect of their jobs. That's one main advantage. Similarly, both go and chess took large initiatives to conquer (15-20 people involved). | |
| Mar 13, 2016 at 9:34 | answer | added | joro | timeline score: -2 | |
| Mar 13, 2016 at 9:08 | answer | added | Brendan McKay | timeline score: 61 | |
| Mar 13, 2016 at 8:55 | answer | added | Mikhail Katz | timeline score: -5 | |
| Mar 13, 2016 at 5:39 | comment | added | Douglas Zare | Natural language processing today can involve far more than the digraph frequencies. It is possible to train a generative model that produces text one letter at a time, yet which closes quotes and understands that it might be in the middle of producing a bibliographical reference. It does seem a huge step to go from playing games, where the problem is primarily to estimate the strength of a position (approximate a function from a space of inputs to $[0,1]$) to writing a coherent proof, but it would be easier in some areas of mathematics than others. Undegraduate real analysis might be easy. | |
| Mar 13, 2016 at 0:50 | comment | added | Derek Elkins left SE | The (potential, and sometimes actual) advantage humans have over machines is that humans can use machines. The (probably somewhat distant for the "typical" mathematician) future of mathematics is human-computer hybrid approaches. Modern proof assistants are already this: the powerful searching abilities of a computer guided by a human. | |
| Mar 13, 2016 at 0:48 | comment | added | Joseph O'Rourke | "a machine learning algorithm trained on a large database of formal proofs" is not comparable to, say, training an algorithm to learn word-adjacency frequencies (which is well-achieved today). A proof has intricate logical internal structure, made explicit, e.g., by Georges Gonthier's computer proof of the $4$-color theorem. It is not clear that any machine-learning techniques could approach this logical complexity. | |
| Mar 13, 2016 at 0:36 | review | Close votes | |||
| Mar 13, 2016 at 14:16 | |||||
| Mar 13, 2016 at 0:30 | comment | added | Noah Schweber | I strongly disagree with the statement "We know that automated theorem proving is in general impossible." While it is true that efficient (e.g. polynomial-time) theorem-proving is impossible, that's not nearly the same thing. | |
| Mar 13, 2016 at 0:23 | comment | added | Burak | I don't understand the question if you are willing to make so strong assumptions regarding the issues related to computational complexity. Humans are bound to work with recursively enumerable sets of axioms, unless you somehow disprove the Church-Turing thesis. The set of theorems that can be proven from an r.e. axiom set is r.e and hence a computer is capable of proving any theorem that humans can prove, if it is provided sufficient resources. | |
| Mar 13, 2016 at 0:13 | history | asked | Māris Ozols | CC BY-SA 3.0 |