Timeline for Can AI make aesthetic judgments?
Current License: CC BY-SA 4.0
16 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Sep 19 at 18:47 | comment | added | 8Mad0Manc8 | AI makes prediction upon the patterns in the data it receives if its predictions are reinforced by the subject of that data in response to that prediction then that AI has knowledge of patterns that please you. It cannot determine why those patterns please you, only they do. Just like any other human cannot appreciate your perspective of some art unless they share that appreciation. . | |
| Sep 16 at 23:00 | answer | added | user71399 | timeline score: 1 | |
| Sep 4, 2023 at 17:13 | comment | added | Christian Hennig | A key issue here is what we (as "observers of the AI") accept as valid aesthetic judgement. The core of the question "can AI do this or that?" is very often a question for conditions that we as observers would require to attribute "doing this or that" to AI. | |
| Sep 4, 2023 at 14:07 | comment | added | Dave | @forlove1 I believe this question could be improved by indicating which/whose conception of "aesthethic judgements" is the context for the question. | |
| Sep 3, 2023 at 19:15 | comment | added | Barmar | BTW, there's an article in this month's Scientific American about an AI that was trained to predict hit songs based on monitoring physiological reactions of listeners, which has a good track record. But "hit" is not the same as "good", and also it's not judging the music itself but the way human listeners react. AIs that were trained to pick hits based on acoustical attributes were not successful. | |
| Sep 3, 2023 at 19:10 | comment | added | Barmar | Even if it did, how would you know if they were "correct"? Esthetics are subjective, and it's rare that there's agreement on what's good. "It's a Wonderful Life" and "The Wizard of Oz" were both flops when they first came out, but are now considered beloved classics. "Citizen Kane" is considered by most critics to be the best American film in history, but lovers of modern action movies would considered it completely boring. | |
| Sep 3, 2023 at 15:28 | answer | added | Angjelin Hila | timeline score: 1 | |
| Sep 3, 2023 at 13:50 | history | became hot network question | |||
| Sep 3, 2023 at 13:26 | answer | added | Jani Miettinen | timeline score: 0 | |
| Sep 3, 2023 at 12:17 | answer | added | Professor Sushing | timeline score: 2 | |
| Sep 3, 2023 at 8:02 | comment | added | user67521 | ha confusing as ever @AgentSmith what do you mean?? | |
| Sep 3, 2023 at 7:54 | comment | added | Hudjefa | Set theorectical concepts seem relevant. | |
| Sep 3, 2023 at 6:31 | comment | added | user67521 | i just think that a truly revolutionary moment in art @Conifold would bear the successes and failures of past art, i.e. understand them - in its code - as artworks. i suppose not, for the reasons i have outlined in my answer. cheers | |
| Sep 3, 2023 at 6:29 | comment | added | Conifold | It all depends on what is meant by "aesthetic judgment", which is highly controversial. Can AI make any judgments? If it just means producing up or down verdict on something then sure. And it can do that on art too, just give it a training set of 'beautiful' art. Midjourney already makes new 'art' based on that much. Kant says that "aesthetic judgment" is one based on feelings. If so, then dogs can make them. Can AI? Probably not on the current meaning of "feeling". But that can change, and we can certainly make AI talk about "their feelings". But how is this supposed to revolutionize art? | |
| Sep 3, 2023 at 5:59 | answer | added | user67521 | timeline score: 0 | |
| Sep 3, 2023 at 5:49 | history | asked | user67521 | CC BY-SA 4.0 |