Daniel Buschek

📄 Preprint: arxiv.org/abs/2509.11826

Key findings (14 teams, 1 week): 📍Profiles were seen as personal territory but it's ok to interact with others' agents/tasks/comments 📍User-initative preferred 📍Thus, teams incorporated agents into social collaboration norms, rather than treating them as "equal" team members.

Typical chatbots force co-writers to leave shared docs. Our #CHI2026 paper explores collaborative AI use in shared docs via 3 features: 🤖 Shared agent profiles ☑️ Repeatable tasks, triggered by users or system 💬 Agents respond in shared comments Preprint in 🧵 w @flolehmann_de

@PreranaKh2 @sensifylab Yes! :)

@DBuschek @sensifylab Hi Prof. Daniel! Do you mean the “When AI Rewrites the News” paper?

Late post, but academia runs on delayed gratification 🙂 Our paper When AI Rewrites the News is accepted to #CHI2026 🎉 📍 Barcelona | April 2026 🇪🇸 Grateful to Dr. Benjamin Bagozzi, Varun Pappu, and Matthew Louis Mauriello. @sensifylab #HCI #AIandMedia

@eddiejiao_obj Really cool "semantic resize"! :) We're also working on some more canvas stuff.

@DBuschek very cool! was playing with some text direct manipulation earlier x.com/eddiejiao_obj/…

📱How might mobile text UIs better integrate LLMs? We designed “spread-to-generate”: Spreading fingers on a draft, bubbles show expected length, then get filled with generated words. This achieves continuous control UX despite LLM token streaming latency. #CHI2025 preprint👇

@eddiejiao_obj You might also like this new project by Xiaohan Peng et al.: youtube.com/watch?v=3lxWq0…

@eddiejiao_obj Agree - there's still a large unexplored design space for interaction with generative systems. Great work with your stylus-based concept! In case you're interested, we explored pinch/zoom for controlling text generation - maybe sth for your text cards 😄 x.com/DBuschek/statu…

the people yearn for direct manipulation!

Writing an HCI paper about an AI-powered system to a venue like UIST 2026 or CHI 2027? Wondering what reviewers expect you to report, and how to approach paper framing and writing? Check out our reporting guidelines: medium.com/p/7c3ae86341e3…

@jocelynjshen Great work! We explored direct manipulation via touch interaction with LLMs at CHI last year as well, including a design space and pinch/zoom gestures. See: x.com/DBuschek/statu… Or dl.acm.org/doi/10.1145/37…

Excited to share our #CHI2026 paper “Texterial: A Text-as-Material Interaction Paradigm for LLM-Mediated Writing” (done during internship at Microsoft Research) We imagine interacting with LLMs by treating text as a material like plants/clay. 📃arxiv.org/pdf/2603.00452 🧵[1/n]

📢 Looking for current research on #HCI + #AI? Here's a categorised collection of 300+ #CHI2026 preprints, collected via arXiv: dbuschek.medium.com/chi26-preprint…

@IanArawjo Almost had to cancel a small practical course because of a lack of very basic programming skills. 😐 String concat was also a question. It's a study module where students can choose, i.e. they explicitly signed up for that course. So I think, some "don't know they don't know."

I encountered this too. One didn’t know how to do string concatenation in JavaScript. String concatenation. Literally a +. Over-reliance is becoming a major issue—people are forgetting, or not even learning, the very basics of programming.

@MattNiessner Unfortunately AI "polish" can reduce quality if it ends up as more than polish: Replacing the thinking that would make those paper parts valuable (regardless of whether they land in perfect English or not), e.g. synthesis of related work. Wrote on it here: @dbuschek/when-llms-write-our-papers-1cc746373cd0" target="_blank" rel="nofollow noopener">medium.com/@dbuschek/when…

Historically, academia used presentation quality as a proxy for scientific merit. Now that AI is eliminating polish overhead, everyone is confused, often stuck in debates whether we should allow LLMs. On the bright side, we are finally forced to evaluate the actual research content rather than extrapolating value from the text and visuals.

@thomasmahler Is this a rune attack that consumes a fish instead of focus? 😁

Just saw this in our commit list. Hold on to your butts, Fish Slap is coming! 😂🤣

@ThePhDPlace @reviewer3com Interesting - self-checking before submission seems useful! I'm worried about the use as a shortcut though, to just get a review done without caring for quality at all. I've written about my anecdotal experiences with this here: medium.com/p/486028b01668

Thanks to @Reviewer3com for letting me take a look and share how it works. If you want to try it yourself, it’s reassuring to know that everything runs privately, and your paper isn’t stored or shared. You can try it here: reviewer3.com

More than half of researchers now use AI in peer review, according to Nature. The question is whether AI can actually doanything useful in that process, or whether it just suggests adding a few emdashes… So I tested it. Here’s what I found 👇 #ad

Reminder for #CHI2026 ACs: You can still click "excellent review" (for your 1AC papers) to award specific reviewers for great work!

Finalising your #CHI2026 revision? Or suffering from not being allowed to do so? I've "reviewed-to-reject" one of our papers with ChatGPT, to show how this leads to bad reviews - and give ideas for responding to this: @dbuschek/dont-review-with-an-llm-laundry-list-method-486028b01668" target="_blank" rel="nofollow noopener">medium.com/@dbuschek/dont…

@IanArawjo Interesting! We might also need more structured reviewing formats. People can use AI to generate an infinite number of issues for any paper, whether adequate for that paper's methodology or not - while authors only get 5k characters to defend their work against that nonsense.

HCI venues need to seriously consider making their review processes more public than they are now. Going forward, less transparency may only hinder our ability to perform similar meta-analyses to diagnose problems as peer review demands expand.

@IanArawjo Make that 3 papers. Also, we've submitted both a quant and a qual paper that were each met with critique based on expecting it to be more on the other end. 🙄 Hence my earlier post about reviewing-to-reject.

The fact that I have to, in 2 separate papers, tell readers what positivism and interpretivism is to nudge them that they are being too positivist in their expectations, is a big problem for HCI. One would have thought that Crabtree and Soden et al.'s rejoinders would be enough.

Whether you're new or a seasoned reviewer, I highly recommend Ken Hinckley's article on excellence in reviews - and championing papers: kenhinckley.wordpress.com/wp-content/upl…

Have we lost peer appreciation?🤔Not just at #CHI2026 I notice many review-to-reject. All studies have tradeoffs & limited scope. Sure, reviews are fast even with "passing knowledge", if we list what's not there to conclude it's lacking. But we should review what *has* been done.