Artificial intelligence and illusions of understanding in scientific research

Messeri, Lisa; Crockett, M. J.

doi:10.1038/s41586-024-07146-0

Cited by 49 publications

(14 citation statements)

References 160 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…AIPAs may inadvertently reinforce users’ knowledge bubbles, 4 potentially deprioritizing environmentally relevant topics if they are not already part of the user’s interests. The companies supplying these assistants may also have their own agendas, which could influence the information provided to users.…”

Section: Potential Risksmentioning

confidence: 99%

AI Personal Assistants and Sustainability: Risks and Opportunities

Rillig,

Kasirzadeh

2024

Environ. Sci. Technol.

View full text Add to dashboard Cite

I n the near future, households and businesses may increasingly rely on advanced AI personal assistants (AIPAs) to manage their daily operations, with potentially transformative effects on energy use, carbon emissions, and sustainability. AIPAs are digital agents powered by generative AI, such as large language models (LLMs) like those that underpin ChatGPT, which can produce human-like text, image, and video. 1 These AI-driven assistants interact with users through natural language interfaces, translating human requests into computer commands to manage various devices, tasks, and applications. 2 Current AIPAs, such as Amazon's Alexa, Microsoft's Cortana, Google Assistant, or Apple's Siri, are designed to assist with a relatively narrow set of tasks, and as such, they are far cry from Joi, the holographic and immensely capable AIPA featured in the movie "Blade Runner 2049". However, as the rapid technological development in generative AI continues, future AIPAs may become capable of mastering highly complex tasks that involve content creation, device integration, communication, coaching, real-time information retrieval and

show abstract

Section: Potential Risksmentioning

confidence: 99%

AI Personal Assistants and Sustainability: Risks and Opportunities

Rillig,

Kasirzadeh

2024

Environ. Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…[9][10][11] For manuscript preparation, researchers have used GAI tools to conduct literature review, formulate study design, analyze data, and even interpret results. [12][13][14][15] For peer review, GAI has been used to summarize manuscript contents, review code, check methods, and even draft comments and feedback. 9,16 These practices highlight that such tools have and will continue to transform scholarly publishing.…”

Section: Introductionmentioning

confidence: 99%

A Systematic Examination of Generative Artificial Intelligence (GAI) Usage Guidelines for Scholarly Publishing in Medical Journals

Yin,

Lu,

et al. 2024

Preprint

View full text Add to dashboard Cite

Background A thorough and in-depth examination of generative artificial intelligence (GAI) usage guidelines in medical journals will inform potential gaps and promote proper GAI usage in scholarly publishing. This study aims to examine the provision and specificity of GAI usage guidelines and their relationships with journal characteristics. Methods From the SCImago Journal Rank (SJR) list for medicine in 2022, we selected 98 journals as top journals to represent highly indexed journals and 144 as whole-spectrum sample journals to represent all medical journals. We examined their GAI usage guidelines for scholarly publishing between December 2023 and January 2024. Results Compared to whole-spectrum sample journals, the top journals were more likely to provide author guidelines (64.3% vs. 27.8%) and reviewer guidelines (11.2% vs. 0.0%) as well as refer to external guidelines (85.7% vs 74.3%). Probit models showed that SJR score or region was not associated with the provision of these guidelines among top journals. However, among whole-spectrum sample journals, SJR score was positively associated with the provision of author guidelines (0.85, 95% CI 0.49 to 1.25) and references to external guidelines (2.01, 95% CI 1.24 to 3.65). Liner models showed that SJR score was positively associated with the specificity level of author and reviewer guidelines among whole-spectrum sample journals (1.21, 95% CI 0.72 to 1.70), and no such pattern was observed among top journals. Conclusions The provision of GAI usage guidelines is limited across medical journals, especially for reviewer guidelines. The lack of specificity and consistency in existing guidelines highlights areas deserving improvement. These findings suggest that immediate attention is needed to guide GAI usage in scholarly publishing in medical journals.

show abstract

“…Doctor-patient communication is the key to physician-patient relationship mitigation, and awareness and cognition play a crucial role in doctor-patient communication [25]. Consciousness is our ability to perceive our inner thoughts and the external world, enabling us to understand and interpret our experiences while also recognizing the existence and state of others [26,27].…”

Section: Introductionmentioning

confidence: 99%

The DIKWP (Data, Information, Knowledge, Wisdom, Purpose) Revolution: A New Horizon in Medical Dispute Resolution

Mei,

Duan

2024

Applied Sciences

View full text Add to dashboard Cite

The doctor–patient relationship has received widespread attention as a significant global issue affecting people’s livelihoods. In clinical practice within the medical field, applying existing artificial intelligence (AI) technology presents issues such as uncontrollability, inconsistency, and lack of self-explanation capabilities, even raising concerns about ethics and morality. To address the problem of doctor–patient interaction differences arising from the doctor–patient diagnosis and treatment, we collected the textual content of doctor–patient dialogues in outpatient clinics of local first-class hospitals. We utilized case scenario analysis, starting from two specific cases: multi-patient visits with the same doctor and multi-doctor interaction differences with the same patient. By capturing the external interactions and the internal thought processes, we unify the external expressions and internal subjective cognition in doctor–patient interactions into interactions between data, information, knowledge, wisdom, and purpose (DIKWP) models. We propose a DIKWP semantic model for the doctor–patient interactions on both sides, including a DIKWP content model and a DIKWP cognitive model, to achieve transparency throughout the entire doctor–patient interaction process. We semantically–bidirectionally map the diagnostic discrepancy space to DIKWP uncertainty and utilize a purpose-driven DIKWP semantic fusion transformation technique to disambiguate the uncertainty problem. Finally, we select four traditional methods for qualitative and quantitative comparison with our proposed method. The results show that our method performs better in content and uncertainty handling. Overall, our proposed DIKWP semantic model for doctor–patient interaction processing breaks through the uncertainty limitations of natural language semantics in terms of interpretability, enhancing the transparency and interpretability of the medical process. It will help bridge the cognitive gap between doctors and patients, easing medical disputes.

show abstract

Artificial intelligence and illusions of understanding in scientific research

Cited by 49 publications

References 160 publications

AI Personal Assistants and Sustainability: Risks and Opportunities

AI Personal Assistants and Sustainability: Risks and Opportunities

A Systematic Examination of Generative Artificial Intelligence (GAI) Usage Guidelines for Scholarly Publishing in Medical Journals

The DIKWP (Data, Information, Knowledge, Wisdom, Purpose) Revolution: A New Horizon in Medical Dispute Resolution

Contact Info

Product

Resources

About