Anil Doshi @anilrdoshi.com · 25/02/2026This thread has brought me some sense of relief for the same thought I had recurring for the past 10(!) years 010
Anil Doshi @anilrdoshi.com · 23/02/2026My Tedx talk to high school students about a year ago was very much along these lines: m.youtube.com/watch?v=WRMY...m.youtube.comHow to use AI without killing your creativity | Dr Anil Doshi | TEDxWestminster SchoolYouTube video by TEDx Talks 000
Reposted by Anil DoshiRené Walter @rawx.bsky.social · 15/01/2026"Analysis of 41 million papers finds that although AI expands individual impact, it narrows collective scientific exploration". This is the science-equivalent to "Generative AI enhances individual creativity but reduces the collective diversity of novel content" www.science.org/doi/10.1126/...science.orgGenerative AI enhances individual creativity but reduces the collective diversity of novel contentGenerative AI can enhance the creativity of short stories but may limit the variation in diverse outputs. 32311
Anil Doshi @anilrdoshi.com · 15/01/2026My co-author, Oliver, and I had a similar thought. It is validating to see our result appear in different contexts with different groups. Thanks for pointing this out to the Bsky community! 020
Anil Doshi @anilrdoshi.com · 13/01/2026Or read the human–AI task tensor book chapter: papers.ssrn.com/sol3/papers....papers.ssrn.comToward a Human–AI Task Tensor: A Taxonomy for Organizing Work in the Age of Generative AIWe introduce a framework for understanding the impact of generative AI on human work, which we call the human–AI task tensor. A tensor is a structured framework 000
Anil Doshi @anilrdoshi.com · 13/01/2026Third dimension of the human–AI task tensor is AI modality: whether the AI is embodied digitally or physically. Read more in my substack post here: open.substack.com/pub/dosh/p/t...open.substack.comThe third dimension of the Human–AI task tensor: interaction modality takes AI from the screen to the streetThe medium of interaction shapes how humans and AI work together and how much autonomy AI might be able to make. 100
Anil Doshi @anilrdoshi.com · 12/11/2025Pretty amazing tool. Did you have to sort permissions for searching full text of the journal papers? 010
Anil Doshi @anilrdoshi.com · 06/10/2025Hard to know. It sometimes boosts outputs from non-experts and sometimes fails to boost output from experts. I think use case is going to matter a lot here. 100
Anil Doshi @anilrdoshi.com · 23/09/2025I posted about a recently published papet in @gjrietveld.bsky.social’s excellent Platform Papers blog. See my post, the original paper, and take a look at Joost’s amazing platform papers Substack and papers database. 020
Anil Doshi @anilrdoshi.com · 04/09/2025Is there a conference schedule to see the presentations? Sounds fascinating. 100
Anil Doshi @anilrdoshi.com · 17/07/2025The second post (just published) covers the first dimension of the tensor: task definition: dosh.substack.com/p/the-first-...dosh.substack.comThe first dimension of the Human-AI task tensor: task definitionConsidering how the clarity and structure of a task shapes effective human-AI collaboration 000
Anil Doshi @anilrdoshi.com · 17/07/2025The first substack post provides an overview of the tensor's eight dimensions: dosh.substack.com/p/mapping-th...dosh.substack.comMapping the Future of Work: An Introduction to the Human–AI Task TensorBook chapter to be published in the Handbook of AI and Strategy edited by Felipe Csaszar and Nan Jia 100
Anil Doshi @anilrdoshi.com · 17/07/2025I have started a series of posts based on a forthcoming book chapter I have published with Alastair Moore titled "Toward a Human–AI Task Tensor: A Taxonomy for Organizing Work in the Age of Generative AI" that I wanted to share. The book chapter is here: papers.ssrn.com/sol3/papers....papers.ssrn.comToward a Human–AI Task Tensor: A Taxonomy for Organizing Work in the Age of Generative AIWe introduce a framework for understanding the impact of generative AI on human work, which we call the human–AI task tensor. A tensor is a structured framework 110
Anil Doshi @anilrdoshi.com · 24/05/2025When I was young, I dreamed of being on Nickelodeon's Double Dare. It would have been the crowning achievement of my young life. 000
Anil Doshi @anilrdoshi.com · 18/03/2025My latest post about the downsides of generative AI. As a reminder to myself not to turn a blind eye to these downsides. open.substack.com/pub/dosh/p/b...open.substack.comBad actors and generative AICombining my research interests on fake news and generative AI suggests an all too easy pathway for malicious actors to do bad things with generative AI 011
Anil Doshi @anilrdoshi.com · 12/03/2025Thanks for sharing Karthik! We are long overdue for a catchup 010
Anil Doshi @anilrdoshi.com · 10/03/2025New post summarizing points I made about generative AI and learning for a talk I gave at TEDxWestminster School. open.substack.com/pub/dosh/p/s...open.substack.comSumming up my thoughts on generative AI and learning in a 10 minute TEDx talkFollowing up on my previous post about generative AI and the classroom, I (along with thousands of other teachers and academic) continue to think about how generative AI should play a role in the clas... 010
Anil Doshi @anilrdoshi.com · 08/01/2025In the paper we investigate why researchers with different levels of experience had such different reactions to AI assistance with their proposals. 000
Anil Doshi @anilrdoshi.com · 08/01/2025When they received generative AI ideas, less experienced researchers view their work as more novel and more impactful (than less experienced researchers who did not get AI ideas) and they viewed their overall research agenda more positively. More experienced researchers felt the exact opposite. 100
Anil Doshi @anilrdoshi.com · 08/01/2025What we find: Researchers with different levels of experience react very differently to generative AI. 100
Anil Doshi @anilrdoshi.com · 08/01/2025The experiment: We randomly assigned wither researchers received generative AI ideas based on their prior work. Outcomes: We ask the researchers to assess the prospective novelty and impact of their research and assess the prospects of their overall research agenda. 100
Anil Doshi @anilrdoshi.com · 08/01/2025The design: We asked 310 researchers (from a wide variety of academic fields) to provide the title and abstract from a recently published article. And then we asked them to provide a new research proposal based on their own work. 100
Anil Doshi @anilrdoshi.com · 08/01/2025New working paper on generative AI and research. My co-authors, Sen Chai and Matthias Tröbinger, and I have a new working paper on how researchers perceive their own research proposals and agenda when exposed to generative AI ideas: papers.ssrn.com/sol3/papers....papers.ssrn.comHow Experience Moderates the Impact of Generative AI Ideas on the Research ProcessAt the heart of scientific discovery are expert researchers who identify research ideas worthy of inquiry. While generative artificial intelligence (AI) technol 130
Anil Doshi @anilrdoshi.com · 06/01/2025The consortium will be held at UCL School of Management in London on 20 May. It is co-organized by me, Oliver Hauser (Exeter), Hila Lifshitz (Warwick), and Pinar Ozcan (Oxford). Please share with your networks/students and encourage them to apply! 000
Anil Doshi @anilrdoshi.com · 06/01/2025For #PhD students researching generative AI in management and adjacent fields: the AI Plus Management Doctoral Consortium is now accepting applications (deadline 31 Jan). There will be opportunities to present and share research for students at all stages. For more info: www.aiplusmanagement.comaiplusmanagement.comHome20 May 2025 UCL School of Management, London UK 153
Anil Doshi @anilrdoshi.com · 23/12/2024Details: 20 May 2025 at UCL School of Management in London. The purpose is to bring together doctoral students conducting work at the intersection of generative AI and management. We welcome applications from students from different disciplines and at various stages in their career. 100
Anil Doshi @anilrdoshi.com · 23/12/2024Applications for the 2025 AI Plus Management Doctoral Consortium are now open (deadline 31 Jan). Please see the website for more information and instructions for applying. Co-organized with Oliver Hauser (Exeter), Hila Lifshitz (Warwick), and Pinar Ozcan (Oxford). www.aiplusmanagement.comaiplusmanagement.comHome20 May 2025 UCL School of Management, London UK 122
Anil Doshi @anilrdoshi.com · 29/11/2024I reckon this is something that LLMs can handle pretty easily for someone who is not comfortable with coding. 120
Anil Doshi @anilrdoshi.com · 29/11/2024Ha! I’m just inclined to let it go. But I can’t bring myself to do it. 010
Anil Doshi @anilrdoshi.com · 29/11/2024I don’t think I want them here. Just trying to figure out if I should fully “abandon” my Twitter account and delete everything. 000
Anil Doshi @anilrdoshi.com · 29/11/2024What are folks doing with their historical tweets and accounts on that other platform? I just stopped using it but considering deleting all my prior content. Thoughts? #ImOverTwitter 310
Anil Doshi @anilrdoshi.com · 28/11/2024This year only student presentations. But here’s to hoping that this might turn into something bigger in future years! You are obviously invited to attend of course! 030
Anil Doshi @anilrdoshi.com · 28/11/2024The event has a lot of great co-organizers from across the UK: Oliver Hauser at @uniofexeter.bsky.social, Hila Lifshitz at University of Warwick, and Pinar Ozcan at @ox.ac.uk. (I have to get them on BlueSky!). 000
Anil Doshi @anilrdoshi.com · 28/11/2024Management researchers and doctoral students doing work on generative AI: mark your calendar! I am very pleased to announce that we will be running the AI plus Management Doctoral Consortium to take place on 20 May 2025 at the @uclofficial.bsky.social School of Management here in London. 393
Anil Doshi @anilrdoshi.com · 18/11/2024Link to a Substack post about the article if you want to read more details about the study. open.substack.com/pub/dosh/p/n...open.substack.comNew Paper Published on Generative AI and Strategic DecisionsComparing the evaluations of AI evaluators with those of human experts in a new paper published in Strategic Management Journal co-authored with Jason Bell, Emil Mirzayev, and Bart Vanneste 000
Anil Doshi @anilrdoshi.com · 18/11/2024The results show the potential for generative AI to be used as a source of feedback and insight into decisions that are fraught with uncertainty (as is often the case with strategic decisions!). 100
Anil Doshi @anilrdoshi.com · 18/11/2024The two studies's results are consistent. (1) individual AI evaluations tend to be inconsistent (when you ask it to pick A vs B, and then B vs A, LLMs often do not pick the same model). (2) aggregations of AI evaluations start to resemble the evaluations made by the human experts. 100
Anil Doshi @anilrdoshi.com · 18/11/2024In 2 studies, we compare how evaluations of business models made by a variety of LLMs compare to those made by human experts. 100
Anil Doshi @anilrdoshi.com · 18/11/2024My co-authors and I published a new paper titled "Generative artificial intelligence and evaluating strategic decisions" in Strategic Management Journal. onlinelibrary.wiley.com/doi/10.1002/...onlinelibrary.wiley.comGenerative artificial intelligence and evaluating strategic decisionsResearch Summary Strategic decisions are uncertain and often irreversible. Hence, predicting the value of alternatives is important for strategic decision making. We investigate the use of generativ..... 1112
Anil Doshi @anilrdoshi.com · 30/09/2024If you have come across this paper and it has influenced how you have approached generative AI in your workplace or research, can you send me a message (or reply here)? Thanks very much! 000
Anil Doshi @anilrdoshi.com · 30/09/2024My paper with Oliver Hauser on generative AI was published in July and has just gotten over 25,000 downloads! Very proud of our work and the impact it has had. We have spoken with a number of companies and the paper has been cited in policy and company reports. www.science.org/doi/10.1126/...science.orgGenerative AI enhances individual creativity but reduces the collective diversity of novel contentGenerative AI can enhance the creativity of short stories but may limit the variation in diverse outputs. 110
Anil Doshi @anilrdoshi.com · 09/02/2024I suppose another way to answer is that the words, clauses, etc are fitted on past data but how they are then outputted in response to novel inputs (prompts) is not necessarily limited to past data. 000
Anil Doshi @anilrdoshi.com · 09/02/2024If I understand, the answer is I am not sure. I suspect that either it does or it could be instructed to consider future uncertainty in its response. That may increase variance in responses but whether the variance arises from extrapolating beyond the past… I’m not sure. 100
Anil Doshi @anilrdoshi.com · 08/02/2024I hope I did that thread properly in my previous post. Anybody have any tools to simplify that? 000
Anil Doshi @anilrdoshi.com · 08/02/2024There are some additional details in the paper itself. We are really excited about these findings and we think there are multiple immediate and practical implications from the work. Please let us know if you have any questions! 200
Anil Doshi @anilrdoshi.com · 08/02/2024So what are the takeaways? The perspective of generative AI can be useful when you average out the weirdness of any one LLM response. It is useful to treat generative AI, not just as one soundboard or assistant, but a crowd of opinion givers. 100