I play valorant with another professor so that if we talk about grant writing during gaming sessions it can count as work

Wrote a blog post on Splat, our light-weight, free affinity diagramming tool: ianarawjo.medium.com/splat-a-free...

What does a class on software development post LLM contain? Design? Syntax of common langs? Programmer thinking? Environments/libraries/bash? Debugging? References and texts especially welcome

Introducing Splat 🟨📌: Affinity diagramming in a single HTML file. Drag notes and organize ideas on a visual board. With semantic search and an agentic AI assistant. Open source and fully local. Check it out here: github.com/ianarawjo/sp...

The state of programming in 2025. Anyways, off to write a CLAUDE•md…

We're admitting PhD students! (Possible co-PIs include @zittrain.bsky.social or @ianarawjo.bsky.social if interests align.) We're excited about: docs.google.com/document/d/1... Interested? Fill out the linked form & apply to Harvard SEAS (listing me as faculty of interest) by Dec 15th!

I recently rewatched the documentary "Indie Game: The Movie" for an upcoming Triple Click bonus episode (don't miss it!) and it portrays Fez as a game in never-ending development hell, with fans wondering if it'd ever come out, because it had been in production for... 3 years. How quaint!

Big layoffs hitting Amazon as soon as tomorrow, including within the video-game division, sources tell Bloomberg www.bloomberg.com/news/article...

have you ever created a video game mod to make characters "look better" (whatever that means to you)? i want to talk to you for an article i'm writing! get at me via maddymyerswrites at gmail dot com. or repost this if it's not you and you just wanna help me spread the word. thanks <3

Maybe it is just post-deadline malaise, but I am becoming tired (again?) of a lot of work that is done in HCI. Who is the work for? What is the point? I wonder if we would be better off forgetting deadlines and just focusing on providing value to people.

#CHI2026 breaking records with 6.7k paper submissions! It's a lot of papers (last year was ~5k, now we are closing in on ~7k!). I know we all can be grumpy about things in HCI but the field is definitely growing and very much engaging with new ideas! I find it really inspiring.

Today at #uist2025 my student Jeffrey Tao presented our new paper "Answering Developer Questions with Annotated Agent-Discovered Program Traces" While it is too late to catch the talk, you can still find Jeff! I'm happy to make connections over email. Paper: andrewhead.info/assets/pdf/t...

LLM safety work often reasons over high-level policies (be helpful & polite), but must tackle on-the-ground cases (unsolicited money advice when stocks are mentioned). This can feel like driving on an unfamiliar road guided by a generic driver’s manual instead of a map. We introduce: Policy Maps 🗺️

I wrote about Silksong www.bloomberg.com/news/newslet...

I can’t think of any other time in modern video-game history that a game as highly anticipated as Silksong has arrived with no early reviews, impressions, leaks, guides, videos, or even forum teases. Soon, everyone will be playing Silksong for the first time. It’s unprecedented and kinda beautiful

Here's one for all the game devs out there www.bloomberg.com/news/newslet...

BREAKING: Silksong will be out on September 4. Two weeks from today. Really. Often, games that take 7+ years to make are plagued by mismanagement and painful burnout. But for Silksong? Team Cherry was having a blast. They still are. This is their story: www.bloomberg.com/news/newslet...

Earlier this month I had a video call with Team Cherry co-founders Ari Gibson and William Pellen, who talked all about what the last seven years have been like for them. It's a very fun story. Coming Thursday on Bloomberg

🚨Calling all HCI researchers!—Have you authored a paper involving an LLM-integrated system over the past two years? Care to speak about your experiences with the publication process? 📃 Email me (my name at gmail com)—we are running a study around this topic, and would love to hear from you.

Triple Click Live was SO FUN. I couldn't have asked for a better night, and am still emotionally recovering. Thanks to everyone who came out! Let's do it again sometime!! 🎶🎮🎤

HCI scholars -- do you know if there's a mechanism through which one could obtain reviews of HCI papers for qual analysis (post-submit period)? We are performing meta-research and seeing reviews could really help. Is there anything against this, or how would you go about it?

I'm still in shock and so so deeply angry about Vox Media's decision to sell Polygon. I am one of the over 20 people laid off today from the best damn website out there. I'm so proud of the place we built, and the workplace I helped shape. I'm sure I'll have more to say soon, but processing it now

youtube.com/watch?v=Rir-... Interesting and relevant work by @advaitsarkar.bsky.social at alt.CHI! #chi2025

If you have GDC Vault access and you missed it, my talk Babygirls, Daddies, and Himbos: Updated Archetypes for Compelling Characters is up! I'm confident I can inspire you to think differently about writing your game characters, wrapped in a very weird history lesson: gdcvault.com/play/1035113...

The 🕯️ Candle Test: Can LLMs be tricked into changing their answer to a simple riddle? Turns out most top-of-the-line models—from Claude 3.7 to Llama 3.1—fail at this task, with the sole exception of GPT-4o: chainforge.ai/play/?f=65d5...

HCI researchers, are no-shows becoming even more common at your user studies? I have a feeling this is only getting worse—people feeling emboldened to flake or ghost for things they signed up for. Can’t remember it being this bad 8 years ago. Anyone else?

okay I think I've got it

A constructivist's dream: Earlier this week my 9yo asked me: "How do you make a simulation?" Two days later we've got >30 pages of artwork for a "fox simulator" that we're building in Scratch.

Hi folks! We are trying to understand the impact ChainForge has had in the LLMOps space, whether for practical use or just inspiration. If ChainForge has provided you value in any way (even if you dislike it), take a moment to fill out this brief form: forms.gle/n1DpVM7gimmd...

Now that we have vibe coding, we need vibe testing!

A prophecy for programming in the future ---programming has always, and always will be, in a process of change. It is those who embrace the future monstrosity, as Derrida called it, that lead the space. From the end of the paper, "To Write Code," CHI 2020:

For years we have heard about the right to one’s data, to privacy, to repair, etc. It’s time to seriously consider the Right to be Offline. www.thepublicdiscourse.com/2025/01/96846/

Neat visualization that came up in the ARBOR project: this shows DeepSeek "thinking" about a question, and color is the probability that, if it exited thinking, it would give the right answer. (Here yellow means correct.)

BTW: everyone wondering "how will big tech think about education" should read what I consider the modern holy trinity: - Morgan Ames's "The Charisma Machine" - Justin Reich's "Failure to Innovate" - Audrey Watters's "Hack Education" Baumol's Cost Disease is irresistible (in both senses).

Canvas program history

Is anyone actively working on how to show error bars for prompt engineering / model evals …? I’m at a point feel like it would be nice to chat with someone who has thought about this very deeply. Exhausted the usual papers.

URGENT request for help: I need to build a case (soon) that justifies the "broader impact" of my research (yes, this is related to recent events in the US). If you have ever used Chartability or Data Navigator in your work in any capacity, I would love to know! Please email ([email protected]) or dm me.

Gulf of Distraction

Well that's a wrap! The 15-page NSF-NSERC grant I have been working hard on w/ a collaborator is submitted (with pride)! Now I can get back to relaxing and having a life (i.e. helping out and coding on research projects). Grant funding moratorium until June! 🥳

Microsoft's own research confirms something that was already pretty obvious: relying on a text generating machine to come up with answers erodes critical thinking, and is a method favoured by those who never liked doing critical thinking in the first place advait.org/files/lee_20...

E A G L E S Go birds!

Just in case you thought that removing DEI criteria would mean that everyone is competing equally: NIH is removing grad students from underrepresented backgrounds from the applicant pool altogether. Their applications will not be considered. Other students, not from these backgrounds, will be.

Decision tree that has been sent to Program Officers at NSF

🚨BREAKING. From a program officer at the National Science Foundation, a list of keywords that can cause a grant to be pulled. I will be sharing screenshots of these keywords along with a decision tree. Please share widely. This is a crisis for academic freedom & science.

on top of everything else, it is hard to express how much i resent having to think about these people. i just want to read my little books and think my little thoughts and post my little posts.

I used the new citations feature in the Anthropic API to identify a set of supporting facts for each thought in an R1 CoT. I'm surprised at how well it works.

Has anyone done work mapping out models' reasoning chains? Or importantly, what would an informative "map" look like (e.g. tracking of moments where the model fails to remember a tidbit it had generated earlier, or high surprisal jumps)?

Making steady progress on getting rid of DEIA: Democracy, Ethics, Institutions, and Accountability.

Finally got this prototype to work after 1000 years of debugging. Sweet release 🥲 The idea is I dump a bunch of unstructured notes on a topic in, feed that into gpt-4o-mini, ask it to label sections with a set of types – claim, evidence, assumption, etc. – and display them as coloured highlights