Cat Hicks

13h

It's time 💃 for 🕺 board game club! 💃🕺👯‍♂️

12h

On my way to steal your girl (get underestimated at the strategy table and then crush it) 😎

16h

ppl screaming in my replies that AI does nothing while I can only move my fingers to type this post because of how good I was at using AI tbh

Rothaí@rothaiocht.bsky.social

Wait til the markets find out Anthropic’s valuation is based on Claude giving me magic words to make my health insurance pay my claims.

18h

Women just want AI to help with basic household wife tasks (convincing doctors and insurance companies that they need a functioning immune system to live)

nytimes.comOpinion | This Year, Yet Again, I’m Dressing as GrapesEvery Halloween for nearly two decades, I’ve adorned myself with purple (or black) balloons. It’s a lifelong commitment to the bit.

So I have something better in the NYT for you to read than that Marissa Meyer Doing Gender piece. Please read my former roommate Michael's charming piece about how he's dressed as Grapes every year for Halloween for almost twenty years www.nytimes.com/2026/10/25/o...

Somehow this piece is poignant, funny, and slides psychological insight under your ribs so quick you don't catch it, like a very sharp knife. I remember Michael telling me about Grapes our first Halloween together and thinking, this man is a genius

20h

Gift link here! bsky.app/profile/circ...

This post is unavailable.

(sorry I don't have a gift link, but you know, for you nyt people)

The fact that DORA is pausing (stopping? Canceling?) its annual survey is so interesting to me. I wonder if we are moving out of the aggregate Annual Report, Acronym Frameworks era in devex

Orgs probably are comparable in some ways; the problem is that doing that comparison successfully requires way more data and way better operationalizations measured over a way longer time period than we get in most projects.

This is one of my most buzzkill takes as a quant social scientist in software, and one of the ones that people least like to talk to me about, hah, but I think it is quite interesting to think about. Really to do good quant work of any kind you have to find misleading patterns fascinating too.

The fact that DORA is pausing (stopping? Canceling?) its annual survey is so interesting to me. I wonder if we are moving out of the aggregate Annual Report, Acronym Frameworks era in devex

There is a lot of "persona" thinking in how people talk about developer experience, as if we can easily derive clusters, types of teams, types of people. I am too much a psychologist of environments to like that and I think you can force a lot of false "groups" to emerge that way.

Don't know where it'll all go, but I DO think we have an opportunity in software to break out of some of these approaches and take a much more dynamic look at our within-team change, within-person change, and learning journeys within orgs. That's pretty exciting to me.

The fact that DORA is pausing (stopping? Canceling?) its annual survey is so interesting to me. I wonder if we are moving out of the aggregate Annual Report, Acronym Frameworks era in devex

With DX folded into Atlassian and DORA changing and a lot of the old narratives of developer experience getting troubled by the agentic era (e.g., "Activity" benchmarks looking useless as soon as they were developed...), interesting times

johnflournoy.scienceNo Silver Bullets: Why Understanding Software Cycle Time is Messy, Not Magic

I have generally been a proponent, both in my research writing and in my work with client orgs, of focusing mostly on change-over-time-within-your-own-org, in part because I see massive problems with assuming orgs are comparable (see our paper: johnflournoy.science/no-silver-bu...)

The fact that DORA is pausing (stopping? Canceling?) its annual survey is so interesting to me. I wonder if we are moving out of the aggregate Annual Report, Acronym Frameworks era in devex

I connect this a bit to this: bsky.app/profile/grim... There is a thread I keep running into which is this hunger for personalized diagnostics, personalized over time inference

Cat Hicks@grimalkina.bsky.social

There's a lot of people who never made charts with data making Claude charts with data and I'm going to go against the tide and say I like it. I really don't care if it's messy. Lots of my programmer friends are trying to unlock analyzing data and never really have before. I think it's cool

I have made no secret of the fact that I have found the analytic approach in the last few DORA reports I read very confusing. Cluster analyses, constructs without construct definition, and a lot of causal language but also packaged with some unusual operationalizations bsky.app/profile/grim...

Cat Hicks@grimalkina.bsky.social

This is from the 2025 DORA report. As ever, the report is pretty obsessed with forcing clusters to emerge. I am quite skeptical.

the real slop is the people slopping up my social energy tbh

Cat Hicks@grimalkina.bsky.social

you know, the fact that all three of your replies on here are nothing but angry venting at people makes me decide this isn't a conversation that's a real one

I also just find it really cute 😭 it makes me realize how much we all have different technical pieces and divisions cut us off from doing it together. Software engineers valuing what we do is charming to me. Your first crack at observational causal inference and you're psyched?? I know! I get it 🥰

In This House We Believe: - Identity Threat is Fake - Skills Are For Sharing - Work For Me Will Expand The More People See My Area of Work - Active Learning Matters - Slop Charts Are Just Drafts

A lot of software teams I talk to are trying to understand themselves and their own activity over time. There is an aliveness to it that I like. There's a questioning of just being handed metrics when you feel Claude could help you learn to wade through data. I see the mess but I just love it anyway

I'm just a firm believer that the more you get into evidence methods with real data the more you'll think critically about the world. Idk it's an interesting side of all this I don't see anyone pointing out but it strikes me.

I've always wanted us to be able to be just that little bit more "in the weeds" with each other. Honestly I'd rather read a messy but good hearted attempt at analysis that I disagree with but can talk about with a programmer friend than another extravagantly misleading headline from a university

my face right now 😱😭 I don't know why this is making me so emotional (probably because I taught myself out of library books for the for the first seventeen years of my life)

Marsh Gardiner@earth2marsh.com

Of the public libraries I use, LA was the first system to make the POST ebook available (in the last 3-5 weeks, I'd estimate).

can't tell you how many psychology books I read at the library as a teenager trying to figure out the world around me. Used to finish up my terrible little minimum wage job (since I wasn't in school I worked a ton) and just read there for hours.

I personally think if you want to espouse that you have a solution to how to provide personalized education to children for whom chronological age is a bad sorting mechanism you should have to share it in a full policy write up. Like WHAT IS THE SOLUTION.

Every post I've seen on this, this morning, describes things that exist

nature.comChatGPT Health performance in a structured test of triage recommendations - Nature MedicineA stress test of ChatGPT Health triage revealed missed high-risk emergencies and inconsistent activation of suicide-crisis safeguards, raising safety concerns for consumer-scale deployment.

This kind of thing is exactly why @daniloc.xyz and I continue to believe that the best results for patient outcomes is not just going to come from obsession with general frontier models and nothing more

Scott McGrath@smcgrath.phd

Agreed. The ChatGPT health triage study highlighted how much barricading we need to establish to prevent people from going too fast and flying off a cliff. www.nature.com/articles/s41...

The other thing I wish we could admit is that every side of this could in a given case be wrong. Doctors can be wrong. Science can be wrong. Patients can be wrong. AI can be wrong. Whether we get to the best outcome with the information at hand has to be emergent

I don't trust systems that are designed to cater to only one side of that. Maybe it will end up being something we choose, eg AI that can never say a doctor is wrong, but I question where that would lead us.

Scott McGrath@smcgrath.phd

Agreed. The ChatGPT health triage study highlighted how much barricading we need to establish to prevent people from going too fast and flying off a cliff. www.nature.com/articles/s41...

Reconciliation between your perceptions and your account, the science accounts, and the accounts of the medical system, is a vital thing. And we're constantly left in need of it.

Honestly having watched hundreds of my patient friends use AI I don't think they are using it to replace care they are using it to try to create a missing stair that never existed in the first place I don't think products are getting that

Scott McGrath@smcgrath.phd

Agreed. The ChatGPT health triage study highlighted how much barricading we need to establish to prevent people from going too fast and flying off a cliff. www.nature.com/articles/s41...

I see better outcomes when I sit with people as they use Informed Patient when they have a space of drafting different versions of talking about what they think is going on. I think my psychologist, interviewing, perceptions/beliefs researcher side makes this clear to me

Having a "health thing" going on and not figured out means dealing with sand shifting under your feet, looking back and putting a different name on symptoms, things you didn't even realize were symptoms...etc.

Scott McGrath@smcgrath.phd

Agreed. The ChatGPT health triage study highlighted how much barricading we need to establish to prevent people from going too fast and flying off a cliff. www.nature.com/articles/s41...

And why I've been pretty obsessed with anchoring biases as I think about Informed Patient