shanghaishanghai

人工智能追踪他人心理状态能力如何?国际最新研究称与人类相当。中新网北京5月25日电 (记者 孙自法)人工智能可否追踪他人心理状态?能力如何?施普林格·自然旗下专业学术期刊《自然-人类行为》最新一期发表一篇人工智能研究论文显示,在测试追踪他人心理状态能力(也称心智理论,Theory of Mind)的任务中,两类人工智能——基于规则的系统和深度学习模型——已经能够达到与人类相当的准确度。

这项研究由美国麻省理工学院(MIT)的研究人员领导,他们设计了一系列任务来评估人工智能理解他人意图、情绪和信念的能力。研究人员发现,尽管深度学习模型在某些任务上表现优于基于规则的系统,但两者在整体上均能有效地识别和预测人类行为背后的心理状态。

研究人员表示,这一发现表明,人工智能在理解复杂社交互动方面取得了显著进展,这可能对未来的机器人和自动化系统设计产生重要影响。例如,在医疗、教育和社会服务等领域,人工智能能够帮助诊断心理问题,提供个性化的干预措施,甚至改善人际沟通。

然而,专家同时指出,尽管人工智能在追踪心理状态方面取得了进展,但它们仍然缺乏对复杂情感和社交情境的直觉理解,以及人类所特有的同情心和共情能力。因此,人工智能在完全取代人类心理学家和临床医生之前还有很长的路要走。

总的来说,这项研究为人工智能在心理状态追踪领域的潜力提供了新的视角,同时也为人类与机器之间的互动带来了新的思考。随着技术的不断进步,人工智能在未来可能会在更广泛的社会和临床环境中发挥作用,但它们仍需与人类专家紧密合作,以实现最佳的决策和干预。

英语如下:

Title: “AI’s Ability to Track Mental States Nears Human Level”

Keywords: AI psychology, tracking, comparable

Content: Can AI track the mental states of others? A new international study claims it’s nearly as good as humans.

BEIJING, May 25 (Xinhua) — Can artificial intelligence track the mental states of others? And how good is its ability?

A new study published in the latest issue of the academic journal “Nature Human Behaviour” by Springer Nature shows that two types of artificial intelligence — rule-based systems and deep learning models — have reached a comparable accuracy in tasks to track the mental states of others (also known as Theory of Mind).

The research, led by researchers from the Massachusetts Institute of Technology (MIT), designed a series of tasks to evaluate AI’s ability to understand others’ intentions, emotions, and beliefs. The study found that, while deep learning models performed better in certain tasks than rule-based systems, both types of AI effectively identified and predicted the mental states behind human behaviors overall.

The researchers noted that this finding indicates significant progress in AI’s understanding of complex social interactions, which could have important implications for the design of future robots and automated systems. For instance, in fields such as healthcare, education, and social services, AI could assist in diagnosing psychological issues, providing personalized interventions, and even improving interpersonal communication.

However, experts also pointed out that, while AI has made progress in tracking mental states, it still lacks intuitive understanding of complex emotions and social contexts, as well as the empathy and compassion unique to humans. Therefore, AI has a long way to go before it can fully replace human psychologists and clinicians.

In summary, the study provides a new perspective on the potential of AI in tracking mental states and brings new considerations to human-machine interactions. As technology continues to advance, AI may play a more significant role in a wider range of social and clinical environments in the future, but it will still need to work closely with human experts to achieve the best decisions and interventions.

【来源】http://www.chinanews.com/sh/2024/05-25/10223216.shtml

Views: 1

发表回复

您的邮箱地址不会被公开。 必填项已用 * 标注