Dear colleague, it is undisputed that AI can help with the processing of large research datasets. You also cite relevant sources several times (e.g. pp. 108, 110). However, you yourself come to the conclusion that the use of AI will initially take place in the area of repetitive activities that do not require the application of creativity, innovation (p. 108). And it is precisely this connection that I find interesting. Can AI be used in research to produce truly innovative results? Where innovation results from extensive computation in the sense of pattern recognition, this is certainly the case, as can be seen in pharmaceutical research. But is this really innovation (without wanting to define this term comprehensively)? In research, doesn't innovation start with identifying the research gap and defining the resulting research questions? Of course, generative AI will provide answers to such questions. In my experience, however, these do not go beyond a bachelor's level. Qualitative research in particular requires both a fundamental and a deep understanding of the research topic. Without this, I do not believe that innovative, high quality results can be achieved. I see it as an essential task for us professors to raise this awareness in our students.

Using GenAI for the REF is a no-brainer

"There is a growing interest in how Generative AI can be used to support and streamline research assessment processes. Richard Watermeyer and Lawrie Phipps argue the standardisation and formulaic nature of REF assessment, alongside its cost, make it a prime candidate for where generative AI could relieve academic drudge work...

Yet there are reasons why GenAI in REF processes might be resisted, which have to do with the potential harms of outsourcing peer-review as a human activity to machine intelligence, and in corollary, therefore, concerns of bias, inaccuracy and even de-intellectualisation. There may also be quite plausible concerns of how GenAI tools may be misappropriated by institutions in terms of leveraging unfair advantage and in further exacerbating a culture of gamesmanship already endemic to the REF. GenAI might for instance offer competitive advantage for (those institutions that use it explicitly or tacitly in) selecting, against norm-referenced criteria, the most competitive research outputs and impacts for submission. Yet GenAI might also conceivably, with its ongoing optimisation, provide greater rigour and accuracy to what can be habitually subjective determinations of research excellence, particularly in the context of impact evaluation.

Maybe the REF is just the kind of academic drudgery that GenAI’s use may be reasonably justified for and arguably with less concern of ethical breach or even human betrayal. Could it even provide greater credibility to claims of public accountability for the money spent on UK research? Where so much of the REF’s work may be experienced as an unjustifiable tax on universities – at a point of peak financial vulnerability – and unnecessary incursion on researchers’ ever depleting time the human cost of being without GenAI is surely greater, perhaps on this rare occasion, than with..?"

https://blogs.lse.ac.uk/impactofsocialsciences/2025/01/28/using-genai-for-the-ref-is-a-no-brainer/

@ Ljubomir Jacić: To see peer review not as part of the freedom and self-determination of research, but as “academic drudge work”, seems to me to be a sign of decay. In any case, there will then probably be a new academic discipline, namely the optimization of research proposals for evaluation by GenAI...

Dear Researchers, Scientists, Friends,

The presented research questions aptly emphasize the multidimensional impact of generative artificial intelligence on the scientific research process, from the design stage to the development of results, and the text rightly highlights its role in the rapid analysis of literature, identification of trends, and research gaps, which contributes to the formulation of precise questions and hypotheses. Expanding on this, it is worth noting new areas of exploration, such as the use of generative AI models for designing experiments and simulations, automatically generating code for data analysis, visualizing complex research results in an accessible way, and assisting in the process of writing and editing scientific publications. An important direction for further research is also an in-depth analysis of the ethical and methodological challenges associated with the use of AI in science, including issues related to potential biases in training data, the transparency of algorithms, responsibility for generated results, and the impact on the creativity and originality of research. It is crucial to investigate how scientists can effectively collaborate with AI systems while maintaining control over the research process and ensuring scientific rigor. I thank all researchers and scientists for their past contributions to understanding this transformative role of AI in science and gladly invite further collaboration and continuation of the discussion on the optimal and ethical use of artificial intelligence in future research.

Best regards,

Dariusz Prokopowicz

The AI Scientist: Towards Fully Automated Open-Ended Scientific Discovery

"One of the grand challenges of artificial general intelligence is developing agents capable of conducting scientific research and discovering new knowledge. While frontier models have already been used as aids to human scientists, e.g. for brainstorming ideas, writing code, or prediction tasks, they still conduct only a small part of the scientific process. This paper presents the first comprehensive framework for fully automatic scientific discovery, enabling frontier large language models to perform research independently and communicate their findings. We introduce The AI Scientist, which generates novel research ideas, writes code, executes experiments, visualizes results, describes its findings by writing a full scientific paper, and then runs a simulated review process for evaluation. In principle, this process can be repeated to iteratively develop ideas in an open-ended fashion, acting like the human scientific community. We demonstrate its versatility by applying it to three distinct subfields of machine learning: diffusion modeling, transformer-based language modeling, and learning dynamics. Each idea is implemented and developed into a full paper at a cost of less than $15 per paper. To evaluate the generated papers, we design and validate an automated reviewer, which we show achieves near-human performance in evaluating paper scores. The AI Scientist can produce papers that exceed the acceptance threshold at a top machine learning conference as judged by our automated reviewer. This approach signifies the beginning of a new era in scientific discovery in machine learning: bringing the transformative benefits of AI agents to the entire research process of AI itself, and taking us closer to a world where endless affordable creativity and innovation can be unleashed on the world's most challenging problems..."

Preprint The AI Scientist: Towards Fully Automated Open-Ended Scienti...

https://github.com/SakanaAI/AI-Scientist

New Challenges in Scientific Publications: Referencing, Artificial Intelligence and ChatGPT

"The COVID-19 pandemic has led to a surge in scientific publications, some of which have bypassed the usual peer-review processes, leading to an increase in unsupported claims being referenced. Therefore, the need for references in scientific articles is increasingly being questioned. The practice of relying solely on quantitative measures, such as impact factor, is also considered inadequate by many experts. This can lead to researchers choosing research ideas that are likely to generate favourable metrics instead of interesting and important topics. Evaluating the quality and scientific value of articles requires a rethinking of current approaches, with a move away from purely quantitative methods. Artificial intelligence (AI)-based tools are making scientific writing easier and less time-consuming, which is likely to further increase the number of scientific publications, potentially leading to higher quality articles. AI tools for searching, analysing, synthesizing, evaluating and writing scientific literature are increasingly being developed. These tools deeply analyse the content of articles, consider their scientific impact, and prioritize the retrieved literature based on this information, presenting it in simple visual graphs. They also help authors to quickly and easily analyse and synthesize knowledge from the literature, prepare summaries of key information, aid in organizing references, and improve manuscript language. The language model ChatGPT has already greatly changed the way people communicate with computers, bringing it closer to human communication. However, while AI tools are helpful, they must be u​s​ed carefully and ethically. In summary, AI has already changed the way we write articles, and its use in scientific publishing will continue to enhance and streamline the process."

https://www.researchgate.net/publication/371557536_New_Challenges_in_Scientific_Publications_Referencing_Artificial_Intelligence_and_ChatGPT

Towards a science exocortex

"Artificial intelligence (AI) methods are poised to revolutionize intellectual work, with generative AI enabling automation of text analysis, text generation, and simple decision making or reasoning. The impact to science is only just beginning, but the opportunity is significant since scientific research relies fundamentally on extended chains of cognitive work. Here, we review the state of the art in agentic AI systems, and discuss how these methods could be extended to have even greater impact on science. We propose the development of an exocortex, a synthetic extension of a person's cognition. A science exocortex could be designed as a swarm of AI agents, with each agent individually streamlining specific researcher tasks, and whose inter-communication leads to emergent behavior that greatly extend the researcher's cognition and volition."

https://www.researchgate.net/publication/383258904_Towards_a_science_exocortex

AI, originality, and attribution: Researchers’ perspectives on distinguishing contributions

"Artificial intelligence (AI) is increasingly integrated into research, significantly challenging established scholarly norms around originality, contribution, and authorship. While policies are developing, there is a gap in understanding how individual researchers subjectively perceive and navigate these ambiguities in practice, impacting research integrity.

To explore researchers’ perspectives on distinguishing human versus AI contributions, we conducted semi-structured interviews with 18 researchers (PhD student, Postdoctoral Researcher, Faculty) across diverse disciplines (STEM, Social Sciences, Humanities). Data were analyzed via reflexive thematic analysis, informed by Attribution Theory.

Researchers predominantly conceptualize AI as a sophisticated tool requiring significant human direction, rather than a genuine collaborator. To navigate attributional ambiguity, they rely on subjective heuristics – such as “gut feelings” of ownership and using the labor of the research process as a proxy for conceptual contribution. This creates significant ethical tensions and a desire for clearer, more nuanced guidelines.

They face cognitive and practical challenges applying traditional integrity norms to AI-assisted work. Findings highlight the need for critical dialogue, reflective practices, and nuanced guidelines to uphold research integrity and thoughtfully integrate human value with machine capabilities."

https://www.tandfonline.com/doi/full/10.1080/08989621.2025.2536817