Related papers: People cannot distinguish GPT-4 from a human in a Turing test

People cannot distinguish GPT-4 from a human in a Turing test

URL: http://arxiv.org/abs/2405.08007v1
Date: Thu, 9 May 2024 04:14:09 GMT
Title: People cannot distinguish GPT-4 from a human in a Turing test
Authors: Cameron R. Jones, Benjamin K. Bergen,
Abstract summary: GPT-4 was judged to be a human 54% of the time, outperforming ELIZA (22%) but lagging behind actual humans (67%) Results have implications for debates around machine intelligence and, more urgently, suggest that deception by current AI systems may go undetected.
Score: 0.913127392774573
License: http://creativecommons.org/licenses/by/4.0/
Abstract: We evaluated 3 systems (ELIZA, GPT-3.5 and GPT-4) in a randomized, controlled, and preregistered Turing test. Human participants had a 5 minute conversation with either a human or an AI, and judged whether or not they thought their interlocutor was human. GPT-4 was judged to be a human 54% of the time, outperforming ELIZA (22%) but lagging behind actual humans (67%). The results provide the first robust empirical demonstration that any artificial system passes an interactive 2-player Turing test. The results have implications for debates around machine intelligence and, more urgently, suggest that deception by current AI systems may go undetected. Analysis of participants' strategies and reasoning suggests that stylistic and socio-emotional factors play a larger role in passing the Turing test than traditional notions of intelligence.

Related papers

AI-Driven Agents with Prompts Designed for High Agreeableness Increase the Likelihood of Being Mistaken for a Human in the Turing Test [0.0]
GPT agents with varying levels of agreeableness were tested in a Turing Test. All exceeded a 50% confusion rate, with the highly agreeable AI agent surpassing 60%. This agent was also recognized as exhibiting the most human-like traits.
arXiv Detail & Related papers (2024-11-20T23:12:49Z)
Metacognitive Monitoring: A Human Ability Beyond Generative Artificial Intelligence [0.0]
Large language models (LLMs) have shown impressive alignment with human cognitive processes. This study investigates whether ChatGPT possess metacognitive monitoring abilities akin to humans.
arXiv Detail & Related papers (2024-10-17T09:42:30Z)
Human Bias in the Face of AI: The Role of Human Judgement in AI Generated Text Evaluation [48.70176791365903]
This study explores how bias shapes the perception of AI versus human generated content. We investigated how human raters respond to labeled and unlabeled content.
arXiv Detail & Related papers (2024-09-29T04:31:45Z)
Self-Directed Turing Test for Large Language Models [56.64615470513102]
The Turing test examines whether AIs can exhibit human-like behaviour in natural language conversations. Traditional Turing tests adopt a rigid dialogue format where each participant sends only one message each time. This paper proposes the Self-Directed Turing Test, which extends the original test with a burst dialogue format.
arXiv Detail & Related papers (2024-08-19T09:57:28Z)
GPT-4 is judged more human than humans in displaced and inverted Turing tests [0.7437224586066946]
Everyday AI detection requires differentiating between people and AI in online conversations. We measured how well people and large language models can discriminate using two modified versions of the Turing test: inverted and displaced.
arXiv Detail & Related papers (2024-07-11T20:28:24Z)
How Well Can LLMs Echo Us? Evaluating AI Chatbots' Role-Play Ability with ECHO [55.25989137825992]
We introduce ECHO, an evaluative framework inspired by the Turing test. This framework engages the acquaintances of the target individuals to distinguish between human and machine-generated responses. We evaluate three role-playing LLMs using ECHO, with GPT-3.5 and GPT-4 serving as foundational models.
arXiv Detail & Related papers (2024-04-22T08:00:51Z)
Does GPT-4 pass the Turing test? [0.913127392774573]
The best-performing GPT-4 prompt passed in 49.7% of games, outperforming ELIZA (22%) and GPT-3.5 (20%) We argue that the Turing test continues to be relevant as an assessment of naturalistic communication and deception.
arXiv Detail & Related papers (2023-10-31T06:27:52Z)
Sparks of Artificial General Intelligence: Early experiments with GPT-4 [66.1188263570629]
GPT-4, developed by OpenAI, was trained using an unprecedented scale of compute and data. We demonstrate that GPT-4 can solve novel and difficult tasks that span mathematics, coding, vision, medicine, law, psychology and more. We believe GPT-4 could reasonably be viewed as an early (yet still incomplete) version of an artificial general intelligence (AGI) system.
arXiv Detail & Related papers (2023-03-22T16:51:28Z)
Can Machines Imitate Humans? Integrative Turing Tests for Vision and Language Demonstrate a Narrowing Gap [45.6806234490428]
We benchmark current AIs in their abilities to imitate humans in three language tasks and three vision tasks. Experiments involved 549 human agents plus 26 AI agents for dataset creation, and 1,126 human judges plus 10 AI judges. Results reveal that current AIs are not far from being able to impersonate humans in complex language and vision challenges.
arXiv Detail & Related papers (2022-11-23T16:16:52Z)
Joint Inference of States, Robot Knowledge, and Human (False-)Beliefs [90.20235972293801]
Aiming to understand how human (false-temporal)-belief-a core socio-cognitive ability unify-would affect human interactions with robots, this paper proposes to adopt a graphical model to the representation of object states, robot knowledge, and human (false-)beliefs. An inference algorithm is derived to fuse individual pg from all robots across multi-views into a joint pg, which affords more effective reasoning inference capability to overcome the errors originated from a single view.
arXiv Detail & Related papers (2020-04-25T23:02:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.