Talk:Importance

Latest comment: 2 years ago by Phlsph7 in topic Artificial intelligence

For the previous discussions, see Talk:Importance (disambiguation)

Artificial intelligence

edit

Hello SnowFire and thanks for taking a detailed look at the article. I saw that you removed the section on "Artificial intelligence" based on the claim that it overemphasizes the role of "importance" in this field. I think you may be right. But it is also true that importance plays some role in this field. Here are some quotations from the source "The Importance of Importance" by David Waltz: Human intelligence is shaped by what is most important to us -- the things that cause ecstasy, despair, pleasure, pain, and other intense emotions. The ability to separate the important from the unimportant underlies such faculties as attention, focusing, situation and outcome assessment, priority setting, judgment, taste, goal selection, credit assignment, the selection of relevant memories and precedents, and learning from experience. AI has for the most part focused on logic and reasoning in artificial situations where only relevant variables and operators are specified and has paid insufficient attention to processes of reducing the richness and disorganization of the real world to a form where logical reasoning can be applied ... AI has for the most part neglected these sorts of issues. And I think this is a serious problem, because it raises the question about whether AI, if it continues on its current course, is really up to the challenge of fulfilling its stated long term goals ... What humans think is important really ought to have a role in AI—how could it not?

In the light of this, I hope we can restore a modified version of this section. The following proposal makes a few changes to the first sentence, no changes to the second, and completely replaces the third (and thus removes the controversial passages you mentioned):

In the field of artificial intelligence, one challenge for some complex real-world applications is to implement artificial reasoning based on importance. This concerns, for example, the problem of distinguishing important from unimportant shapes in the process of image recognition when trying to simulate a form of perception. According to David Waltz, advances in artificial importance reasoning are necessary to close some of the gaps in fields where artificial intelligence still lacks behind human intelligence.[1][2]

I appreciate you taking the time to help fix this issue. Please let me know if these changes are in your eyes sufficient to properly represent the source or if more changes are needed. Phlsph7 (talk) 04:52, 6 July 2022 (UTC)Reply

@Phlsph7: Hi, sorry about the slow reply! Your ping didn't seem to work - I didn't get a notification. So I only noticed now after the bot update to add reflist talk.
I appreciate you're working in good faith on this one, but if a passage was restored on the topic, I'd really prefer for it to come from a current computer scientist. David Waltz was writing in 1999 with "The Importance of Importance", which was an eternity ago in AI years, and even worse, Waltz was nearly 60 at the time, making me suspicious of even treating him as an authority on 1999-era tech. Computer science has a reputation as a young man's game. It's hard to describe just how dated the state of AI in 1999 compared to now is; it'd be like citing an automobile engineer writing in 1920 about car technology. More generally, I don't get the impression that among current computer scientists, "importance" in the broad sense is even an issue on their minds. It's reaching too far. Right now, AI is struggling to do things like merely distinguish whether that's a pedestrian about to dart out into the street, or an advertisement banner depicting a person on the wall; the second-order question of importance can't even be asked until you're sure whether that person is real or not. But that's just my personal impression of the state of the art in the field. SnowFire (talk) 12:58, 15 September 2022 (UTC)Reply
Thanks for the response, better late than never. I don't fully agree with you but I understand your concerns and I'm no expert in this field. This is probably the least important part of the article so not too much is lost by leaving it out. Maybe we can pick the topic up again if I or someone else stumbles on a more contemporary article on this issue. Phlsph7 (talk) 14:58, 15 September 2022 (UTC)Reply

References

  1. ^ Waltz, David (15 September 1999). "AAAI-98 Presidential Address: The Importance of Importance". AI Magazine. 20 (3): 18–18. doi:10.1609/aimag.v20i3.1464. ISSN 2371-9621.
  2. ^ Zhou, Chuanxin; Sun, Yi; Wang, Degang; Gao, Qi (18 May 2022). "Fed-Fi: Federated Learning Malicious Model Detection Method Based on Feature Importance". Security and Communication Networks. 2022: e7268347. doi:10.1155/2022/7268347. ISSN 1939-0114.{{cite journal}}: CS1 maint: unflagged free DOI (link)