好奇
閱讀設定
(由人工好奇跳轉過嚟)
呢篇文 需要熟悉呢方面嘅人幫手寫。 |
好奇(粵拼:hou3 kei4;拉丁文:cūriōsus)係指想要獲得資訊或者知識而起嘅一系列行為同情緒反應。好奇係一種心理狀態,可以當係情感嘅一種。
定義
[編輯]喺心理學定義上,情感可以泛指啲同動機相關嘅身心狀態,每種情感都會對應某類特定嘅刺激,亦會對應某拃特定嘅身心變化,令個人有動機去做某啲行為[1]。舉例說明,驚係種情感[2]-
而响心理學上,好奇感係畀人噉描述嘅[3]:
- 英文原文:"... a 'manipulatory motive' - that drives organisms to engage in puzzle-solving behavior that involved no tangible reward."[5]
- -粵文翻譯:(好奇係)一股操作性質嘅動力—會引導生物去作出唔會帶嚟咩具體犒賞嘅解謎行為。
- 英文原文:"... a cognitive induced deprivation that arises from the perception of a gap in knowledge and understanding."[6]
- -粵文翻譯:(好奇係)一股受引起嘅認知缺乏感,源於感知到(自己)响知識同理解上有罅隙。
基於呢啲描述,好奇就可以噉想像[7]:Sec. 2:
- 刺激:源於隻動物感知到啲認知上有刺激性嘅事物(睇下面);
- 身心變化:一隻動物覺得好奇嗰陣,唔會好受壓,彷彿好似件事物係一樣犒賞噉[8];
- 動機行為:佢仲會有探索行為,會(例如)行埋引起佢好奇嗰件事物度,想睇清楚啲件事物。
心理學、神經科學同動物行為學等嘅領域都有對好奇感做研究,會想知(例如)一個人覺得好奇嗰陣,佢個腦裏面發生緊啲乜變化[3]。而且人工智能(AI)領域嘅工作者,仲有嘗試教電腦展現出好似人同第啲動物噉嘅好奇行為[7]:Sec. 3。
起因刺激
[編輯]探索行為
[編輯]遊戲設計
[編輯]睇埋:開放世界
AI 研究
[編輯]
強化學習(RL)可以結合人工好奇嚟用:喺廿一世紀初,AI 最大嘅弱點係專化得滯,教一個 AI 幫手睇病,佢唔會識得(例如)做法律相關嘅判斷,但由現實經驗可知,人有能力學完一樣嘢走去學第樣;噉嘅其中一個重要原因係,人具有好奇心-喺手上資訊唔夠嗰陣,人往往會主動噉去搵新嘅資訊吸收;於是有 AI 研究者就提出咗「人工好奇」嘅概念,主張要用電腦模擬人類嘅好奇心,從而教到 AI 唔使吓吓都要由人類畀資訊佢,而係會曉自己搵資訊吸收[9][10]。
對於人工好奇嘅概念,有人就舉咗個噉嘅例子[11]:
- 想像家陣有個人喺一間超市裏面兜圈噉行,想搵菠菜;
- 喺每一步,個人都行經幾排貨架,呢啲貨架上面並冇菠菜;
- 如果個人係跟簡單 RL 行事嘅,佢會一路噉「周圍貨架冇菠菜,冇任何一個選項得到強化」,進入「永世都唔會離開個圈」嘅狀態;
- 但假如個人具有好奇心嘅能力,曉(例如)有動機想探索未行過嘅路線,隨機噉揀條圈以外嘅路線行-就有可能最後搵到菠菜,或者最少脫離「係噉兜圈」嘅狀態。
有研究者指,呢點就係缺乏好奇心嘅 AI 嘅問題所在-冇好奇心嘅智能體,一定要有人畀有用嘅資訊或者環境佢,先會有能力成長,但喺現實,人成日都會面對「周圍環境冇咩有用資訊」噉嘅情況,要自己去搵資訊;而好奇心正正就係能夠「令人自發噉去搵有用資訊」嘅嘢。要達致人工好奇,一段 RL 演算法起碼要有以下嘅嘢[11]:
- 段演算法要有記憶能力,能夠記住過往嘅經驗;
- 段演算法要識得將「而家呢刻觀察到嘅嘢」攞去同記憶入面嘅片段對比;
- 唔淨只取決於一個行動能唔能夠達到目的,仲要或多或少噉取決於「件行為帶嚟嘅觀察有幾新穎」(詳情可以睇返好奇感啲起因)。
... 呀噉。
睇埋
[編輯]文獻
[編輯]- Berlyne, D. E. (1950). Novelty and curiosity as determinants of exploratory behaviour. British journal of psychology, 41(1), 68.
- Burda, Y., Edwards, H., Pathak, D., Storkey, A., Darrell, T., & Efros, A. A. (2018). Large-scale study of curiosity-driven learning (PDF). arXiv preprint arXiv:1808.04355.
- Ciolfi, L., & Bannon, L. (2002, September). Designing Interactive Museum Exhibits: Enhancing visitor curiosity through augmented artefacts. In Eleventh European Conference on Cognitive Ergonomics (Vol. 7).
引咗
[編輯]- ↑ Panksepp, J. (2004). Affective neuroscience: The foundations of human and animal emotions. Oxford university press. p. 9. "Our emotional feelings reflect our ability to subjectively experience certain states of the nervous system."
- ↑ LeDoux, J. E. (2012). Evolution of human emotion: a view through fear. In Progress in brain research (Vol. 195, pp. 431-442). Elsevier.
- ↑ 3.0 3.1 Kidd, C., & Hayden, B. Y. (2015). The psychology and neuroscience of curiosity. Neuron, 88(3), 449-460.
- ↑ James, W. (1983). Talks to Teachers on Psychology and to Students on Some of Life's Ideals (Vol. 12). Harvard University Press.
- ↑ Harlow, H. F., Harlow, M. K., & Meyer, D. R. (1950). Learning motivated by a manipulation drive. Journal of experimental psychology, 40(2), 228.
- ↑ Loewenstein, G. (1994). The psychology of curiosity: A review and reinterpretation. Psychological bulletin, 116(1), 75.
- ↑ 7.0 7.1 Wu, Q., & Miao, C. (2013). Curiosity: From psychology to computation (PDF). ACM Computing Surveys (CSUR), 46(2), 1-26.
- ↑ Costa, Vincent D.; Tran, Valery L.; Turchi, Janita; Averbeck, Bruno B. (2014). "Dopamine modulates novelty seeking behavior during decision making". Behavioral Neuroscience. 128 (4): 556-566.
- ↑ Schmidhuber, J. (2006). Developmental robotics, optimal artificial curiosity, creativity, music, and the fine arts (PDF). Connection Science, 18(2), 173-187.
- ↑ Schmidhuber, J. (2020). Generative adversarial networks are special cases of artificial curiosity (1990) and also closely related to predictability minimization (1991). Neural Networks, 127, 58-66.
- ↑ 11.0 11.1 curiosity artificial intelligence (curiosity AI). Techtarget.
拎
[編輯]- (英文) How can Artificial Intelligence become curious?. Towards Data Science.