iiinova — where your digital self keeps evolving

Carl J. Pearson @carljpearson.bsky.social · 27/08/2026

I’ve decided to get into soccer which is also the first and only sport I’ve ever gotten into. MLS is absolutely getting played with by Liga MX in the Leagues Cup

000

Carl J. Pearson @carljpearson.bsky.social · 27/08/2026

Timely glup shitto reference. Excited to download it today (and look for Sepa)

000

Carl J. Pearson @carljpearson.bsky.social · 27/08/2026

I’ve switched to ChatGPT after a while with Claude. Claude just sounded too “Claude-y” and kept making weird choices. I do really like the UX and integration of the UI in ChatGPT which I didn’t expect.

000

Carl J. Pearson @carljpearson.bsky.social · 24/07/2026

“Genuinely” When it’s not about genuineness whatsoever

100

Carl J. Pearson @carljpearson.bsky.social · 23/07/2026

"Quietly" This has almost no relevant trigger at all, it seems like everything is done "quietly" for claude

100

Carl J. Pearson @carljpearson.bsky.social · 23/07/2026

"XYZ funds ABC" completely out of the context of money or finance.

100

Carl J. Pearson @carljpearson.bsky.social · 23/07/2026

Starting a thread of claude-isms I cannot stand. The equivalent of AI slop artwork in written form #claude

100

Carl J. Pearson @carljpearson.bsky.social · 10/07/2026

interesting that ChatGPT/Codex and Claude Chat/Cowork both collapsed into single interfaces on the exact same update, same day. I have to imagine people are chatting a lot between companies.

000

Carl J. Pearson @carljpearson.bsky.social · 10/07/2026

didn't do much, the LI algo is a mystery to me and now the ads platform is too

000

Carl J. Pearson @carljpearson.bsky.social · 30/06/2026

LinkedIn has absolutely tanked impressions on my posts. Trying a "boost" ad - feels a bit silly but curious to see how it turns out

100

Reposted by Carl J. Pearson

Pavel @spavel.bsky.social · 17/06/2026

"Here's the part most people miss —" ok but what about the part where you shut the fuck up

4494

Carl J. Pearson @carljpearson.bsky.social · 17/06/2026

getting more into codex and cowork - it's crazy how much knowledge you need to piece together on your own to make these tools shine. we really are in the CLI-equivalent days of llm tooling.

000

Carl J. Pearson @carljpearson.bsky.social · 04/06/2026

what's it called when you work on project A, let its prompt run, switch to project B, let its prompt run, switch back to project A after its prompt is done, let the next prompt run, switch to project B after its prompt is done...

000

Carl J. Pearson @carljpearson.bsky.social · 25/05/2026

Running a blog/newsletter stack of WordPress and mailerlite. Is it worth it to switch to ghost? Or is the migration not going to be worth the effort? Definitely not doing substack

110

Carl J. Pearson @carljpearson.bsky.social · 21/05/2026

how did i never see this on a shirt at kubecon

010

Carl J. Pearson @carljpearson.bsky.social · 21/05/2026

Searched a movie on Google I wanted to watch. The AI summary described the entire ending... Guess I'll skip this one

000

Carl J. Pearson @carljpearson.bsky.social · 18/05/2026

I'm pretty sure wearing my 2 month old baby for 3-8 hours of napping a day counts as rucking

000

Carl J. Pearson @carljpearson.bsky.social · 17/05/2026

I just can't deal with anymore Claude-written articles from very smart people I want to hear from. Lack of personality aside, they're also 30-60% too long.

000

Carl J. Pearson @carljpearson.bsky.social · 16/05/2026

I have typically not had bandwidth to cog test even the human translations. I did some light vetting with native speakers who strongly approved of the translations, not even knowing they were AI. I'd like to be more robust but this is working well for now!

010

Carl J. Pearson @carljpearson.bsky.social · 16/05/2026

Semi-related: the most tangible speed up LLMs have had on my #uxresearch process is effectively translating 13 languages in a survey within 10 minutes rather than 10 days.

110

Carl J. Pearson @carljpearson.bsky.social · 16/05/2026

The argument for LLMs being used to help non native speakers is totally valid. But it's two totally separate things to use LLMs to translate non-English text you wrote vs using it to produce the same slop as a native speaker would.

110

Carl J. Pearson @carljpearson.bsky.social · 16/05/2026

IG ads are waaaay better tho, so I'll be back on reels when I need some new clothes.

000

Carl J. Pearson @carljpearson.bsky.social · 16/05/2026

I’ve had more time to rot my brain recently on short form video. It's really interesting how much conservative content reels pushes at me versus tiktok. Definitely some semantic differences in what each prioritizes.

100

Carl J. Pearson @carljpearson.bsky.social · 15/05/2026

It's great having AI for coding but now it feels like a privilege to have learned what code I know pre-AI. It's too tempting to just plug and go without really understanding the mechanics of the models.

000

Carl J. Pearson @carljpearson.bsky.social · 14/05/2026

Working on a big blog post about influence in UX research, defining it, implications for velocity/AI/etc. The big thing that I am feeling is that all roles in a modern product org come back to primarily dealing with political elements of work when you get more senior (IC included).

000

Carl J. Pearson @carljpearson.bsky.social · 13/05/2026

the thing about all of these AI in UXR classes popping up is that I truly believe using AI is not that hard... that was the whole point of AI. Unless you're doing real engineering of the models, you can probably figure out most things on your own without a class.

000

Carl J. Pearson @carljpearson.bsky.social · 09/05/2026

One follow up thought: I always hear "academics often fail in industry because they can't dial down rigor" but never hear those without training fail because they can't dial it up. That's because the quality of rigor is largely invisible to adjacent stakeholders or even UXRs not close to the work.

120

Carl J. Pearson @carljpearson.bsky.social · 08/05/2026

2. There is some random hiccup/latency so songs have a random bar of 3/4 or 5/4 every few songs.

000

Carl J. Pearson @carljpearson.bsky.social · 08/05/2026

1. The clamp force is awful with glasses. Luckily found this gem of a thread with a fix that has people chiming in hundreds of times of 5 years to say how it saved the headphones for them. www.reddit.com/r/airpods/co...

reddit.com

From the airpods community on Reddit

Explore this post and more from the airpods community

100

Carl J. Pearson @carljpearson.bsky.social · 08/05/2026

My Airpod Max 2 headphones are great but...

100

Carl J. Pearson @carljpearson.bsky.social · 01/05/2026

AI for #UXResearch feels like the Microsoft Office of the 2010's. You can put it in your resume, some recruiters expect it, but it's fairly default and doesn't take some magical skill to learn/practice anyway.

000

Carl J. Pearson @carljpearson.bsky.social · 25/04/2026

It's certainly bad for my thinking process but I find I work on two projects at once so I can let Claude or Gemini answer one prompt while I do more work on the secondary project. Would love shorter processing times for the bigger models.

000

Carl J. Pearson @carljpearson.bsky.social · 22/04/2026

A research repo geared towards XFN self serve is so much less powerful than a repo geared towards researcher use + having researchers consistently in the decision-making process.

000

Carl J. Pearson @carljpearson.bsky.social · 21/04/2026

I'm not an AI apologist but I'm curious how much of the "AI is making us dumber" research is similar to Socrates being worried about people's ability to memorize long epics when writing was invented.

010

Carl J. Pearson @carljpearson.bsky.social · 27/02/2026

Can't even go on LinkedIn today with the Block AI Layoff news. If Block had truly found a way to be as efficient with half of their staff, they'd probably keep it quiet and become a consulting company. AI is such a CEO cop out these days 🙄

000

Reposted by Carl J. Pearson

Nicole Bedera @nbedera.bsky.social · 27/01/2026

Those in power have tried to convince us that resistance would be too tiring. But let Minnesotans teach you that it’s actually energizing to be a part of a movement to change society for the better. If you’re tired, get involved. You’ll be surprised by how quickly you start to feel better.

5338103

Reposted by Carl J. Pearson

Margaret Killjoy @margaret.bsky.social · 22/01/2026

I came to Minneapolis to report on what's going on, and one of the main questions I showed up with is "just what is the scale of the resistance?" After all, we're all used to the news calling Portland a "war zone" or whatever when it's just some protests in one part of town.

142153675715

Carl J. Pearson @carljpearson.bsky.social · 11/02/2025

Learning SQL was a simple way to unblock lots of my projects, mainly re: recruiting. As I've specialized more in quant, SQL ability significantly deepens my analysis. All UXRs can benefit from learning SQL. It's about as hard as HTML IMO. carljpearson.com/sql-is-for-a... #uxresearch

carljpearson.com

SQL is for all UX researchers - Carl J. Pearson, PhD

See why UX researchers need to learn SQL and what aspects are most important to focus on.

000

Carl J. Pearson @carljpearson.bsky.social · 11/02/2025

The trick here is not getting drawn in to massive user interview panel companies thinking you're going to get high quality B2B participants from technical fields.

000

Carl J. Pearson @carljpearson.bsky.social · 29/01/2025

Is bluesky where we go to post our UXR hot takes? I haven't used it much but that might motivate me more.

000

Carl J. Pearson @carljpearson.bsky.social · 14/01/2025

I can't tell you how many survey projects I've stopped before they've started due to this calculation: carljpearson.com/the-survey-v... It's not my goal to end survey projects, but being a survey specialist I probably have the opportunity more than the average researcher. #uxresearch

carljpearson.com

The Survey Viability Score: My first question in survey planning

Calculate the survey viability score to see if your survey will reach the right sample size for success - Carl J. Pearson, PhD

041

Carl J. Pearson @carljpearson.bsky.social · 05/12/2024

These same people motivate their push for speed w/ aligning with the "business" needs rather than a vague sense of user-centrism This doesn't make sense because rigor doesn't exist to be performative Rigor exists to find "true" information (whether that is singular or multiple, epistemologically)

000

Carl J. Pearson @carljpearson.bsky.social · 05/12/2024

On LI I see more and more people (often selling consulting) pushing the idea that research needs to be faster, faster, faster. There's a fine line to walk for rigor/speed. If we overstep the line, we can still impact the product, but who knows how the business will actually perform with that intel.

110

Carl J. Pearson @carljpearson.bsky.social · 05/12/2024

Hello! I've had this account for some years now, but it appears we finally have traction to build some community. Find my longer form thoughts here: carljpearson.com/ux-research-...

carljpearson.com

UX Research Blog - Carl J. Pearson, PhD UX Research Blog

Carl J. Pearson's blog about UX research and quantitative methods.

010