Simon P. Couch
LightNews LightNews
banner
simonpcouch.com
Simon P. Couch
@simonpcouch.com
3.5K followers 200 following 230 posts
he/him - writing statistical software at Posit, PBC (née RStudio)🥑

simonpcouch.com, @simonpcouch elsewhere
Posts Replies Media Videos
Reposted by Simon P. Couch
smachlis.bsky.social
Turn Claude into an expert on writing {testthat} #Rstats 📦 version 3 R tests with the Claude Testing R Packages skill by @posit.co
github.com/posit-dev/sk...
That repo has other skills including how to use brand.yml with Quarto and Shiny
github.com/posit-dev/sk...
#GenAI
skills/r-lib/testing-r-packages at main · posit-dev/skills
A collection of Claude Skills from Posit. Contribute to posit-dev/skills development by creating an account on GitHub.
github.com
December 24, 2025 at 12:50 PM
Reposted by Simon P. Couch
colincarlson.bsky.social
This whole thing is amazing, but tucked away in a dropdown at the end of the markdown is... quite seriously the best essay I've read on the environmental impacts of AI?
vruba.bsky.social Charlie Loyd @vruba.bsky.social · 3d
I just published Potato, a new pansharpening package. It aims to render certain kinds of satellite imagery more clearly and accurately than what’s for sale and on satellite maps today: github.com/celoyd/potato/
GitHub - celoyd/potato: A small pansharpening model
A small pansharpening model. Contribute to celoyd/potato development by creating an account on GitHub.
github.com
December 23, 2025 at 10:49 PM
Reposted by Simon P. Couch
grrrck.xyz
Happy holidays! 🎄 I snuck in a {btw} release just before CRAN went on vacation for the year. {btw} is an #RStats package full of tools for LLMs, including a new tool for running R code. Here's a quick demo using `btw_app()` to create a quick festive ggplot2 from holiday movie titles. Cheers! 🧑‍🎄
December 23, 2025 at 4:40 PM
Reposted by Simon P. Couch
maggieappleton.com
Context windows out to thwart all our beautiful dreams
The butterfly meme where the butterfly says "context 95% full. start a new chat" and the guy says "is this AGI?"
December 22, 2025 at 5:23 PM
simonpcouch.com
In this edition of the @posit.co AI Newsletter, some 🎄Year In Review🎄, notes on GPT 5.2, and an experimental AI product for RStudio users.

posit.co/blog/2025-12...
2025-12-19 AI Newsletter - Posit
In this edition of the Posit AI Newsletter, we dive into the most important AI news and research from the past two weeks. Curated for the data science community, we cover everything from model updates...
posit.co
December 19, 2025 at 5:51 PM
simonpcouch.com
infer 1.1.0, a package implementing an expressive grammar for statistical inference, is on CRAN! In a change originally motivated by @allendowney.bsky.social's posit::conf(2024) keynote, the package now supports arbitrary test statistics.

Read more: github.com/tidymodels/i...
A hexagonal logo for "infer" featuring a green Douglas Fir tree centered on a mountain landscape background with blue-gray peaks and white snow-capped ridges. The word "infer" appears in a modern, lowercase black font at the bottom of the hexagon, which has a dark blue border.
December 18, 2025 at 2:27 PM
Reposted by Simon P. Couch
libbyheeren.bsky.social
Why do I drink coffee past noon? Why do I do this thing?
December 17, 2025 at 7:20 PM
Reposted by Simon P. Couch
posit.co
Thinking about running local models with coding agents like Positron Assistant? 💻

A new evaluation by @simonpcouch.com and @sara-altman.bsky.social suggests we aren't there yet. Between hardware hurdles and performance gaps, hosted models are still the best bet for now.

posit.co/blog/local-m...
Local models are not there (yet) - Posit
LLMs that can run on your laptop are not yet capable enough to drive coding agents.
posit.co
December 17, 2025 at 3:14 PM
simonpcouch.com
While coding agents like Positron Assistant and side::kick() are technically compatible with LLMs small enough to run on your laptop, the results leave much to be desired.

New on the @posit.co blog from @sara-altman.bsky.social and I: posit.co/blog/local-m...
Local models are not there (yet) - Posit
LLMs that can run on your laptop are not yet capable enough to drive coding agents.
posit.co
December 16, 2025 at 8:35 PM
Reposted by Simon P. Couch
ctrl-n.bsky.social
Great and easy to understand post on using LLMs, i.e. in #positron #rstats
December 12, 2025 at 3:02 PM
simonpcouch.com
GPT 5.2 is now included in our R code generation eval! The -Pro version is slightly SoTA at a substantially higher price point than similar performers.

Read more: skaltman-model-eval-app.share.connect.posit.cloud
A scatter plot comparing AI models by R coding accuracy (percent correct) versus total cost in USD, with points colored by provider (Anthropic in purple, Google in yellow, OpenAI in blue). Claude Opus 4.5 and GPT-5.2 Pro achieve the highest accuracy (~70%), but Claude Opus 4.5 does so at roughly one-sixth the cost of GPT-5.2 Pro.
December 12, 2025 at 6:20 PM
Reposted by Simon P. Couch
posit.co
New data from @sara-altman.bsky.social & @simonpcouch.com, using the vitals and ellmer packages, looks at LLM performance metrics for #RStats.

👀 Sneak peek: Claude Opus 4.5, Claude Sonnet 4.5, & OpenAI GPT-5 lead in generating correct R code.

Read the full breakdown: posit.co/blog/r-llm-e...
Which LLM writes the best R code? - Posit
The vitals package delivers a dedicated framework for measuring how well Large Language Models perform with code.
posit.co
December 11, 2025 at 3:49 PM
Reposted by Simon P. Couch
rpodcast.bsky.social
Episode 215 of R Weekly Highlights is out! serve.podhome.fm/episodepage/...

🤖 AI Newsl @simonpcouch.com @sara-altman.bsky.social
🎙️ The Test Set @mchow.com @hadley.nz @wesmckinney.com
🤔 Haskell for DS @jonocarroll.fosstodon.org.ap.brid.gy

h/t @mike-thomas.bsky.social Sam Parmar 🙏

#RStats #databs
Issue 2025-W50 Highlights
How the recent frontier LLM model releases compare for successfully generating R code, our take on the new Test Set data science podcast, and a surprising…
serve.podhome.fm
December 10, 2025 at 12:11 PM
simonpcouch.com
A new release of chores is now on #rstats CRAN!

There are now some LLMs tiny enough to run on a laptop that are capable of powering chores helpers. Here's a real-time demo with a model running on my Macbook.

Read more: www.simonpcouch.com/blog/2025-12...
December 10, 2025 at 3:25 PM
Reposted by Simon P. Couch
andrew.heiss.phd
Some closing thoughts for my students this semester on LLMs and learning #rstats datavizf25.classes.andrewheiss.com/news/2025-12...
Will you incorporate LLMs and AI prompting into the course in the future? No. Why won’t you incorporate LLMs and AI prompting into the course? These tools are useful for coding (see this for my personal take on this). However, they’re only useful if you know what you’re doing first. If you skip the learning-the-process-of-writing-code step and just copy/paste output from ChatGPT, you will not learn. You cannot learn. You cannot improve. You will not understand the code. In that post, it warns that you cannot use it as a beginner: …to use Databot effectively and safely, you still need the skills of a data scientist: background and domain knowledge, data analysis expertise, and coding ability. There is no LLM-based shortcut to those skills. You cannot LLM your way into domain knowledge, data analysis expertise, or coding ability. The only way to gain domain knowledge, data analysis expertise, and coding ability is to struggle. To get errors. To google those errors. To look over the documentation. To copy/paste your own code and adapt it for different purposes. To explore messy datasets. To struggle to clean those datasets. To spend an hour looking for a missing comma. This isn’t a form of programming hazing, like “I had to walk to school uphill both ways in the snow and now you must too.” It’s the actual process of learning and growing and developing and improving. You’ve gotta struggle. This Tumblr post puts it well (it’s about art specifically, but it applies to coding and data analysis too): Contrary to popular belief the biggest beginner’s roadblock to art isn’t even technical skill it’s frustration tolerance, especially in the age of social media. It hurts and the frustration is endless but you must build the frustration tolerance equivalent to a roach’s capacity to survive a nuclear explosion. That’s how you build on the technical skill. Throw that “won’t even start because I’m afraid it won’t be perfect” shit out the window. Just do it. Just start. Good luck. (The original post has disappeared, but here’s a reblog.) It’s hard, but struggling is the only way to learn anything. You might not enjoy code as much as Williams does (or I do), but there’s still value in maintaining codings skills as you improve and learn more. You don’t want your skills to atrophy. As I discuss here, when I do use LLMs for coding-related tasks, I purposely throw as much friction into the process as possible: To avoid falling into over-reliance on LLM-assisted code help, I add as much friction into my workflow as possible. I only use GitHub Copilot and Claude in the browser, not through the chat sidebar in Positron or Visual Studio Code. I treat the code it generates like random answers from StackOverflow or blog posts and generally rewrite it completely. I disable the inline LLM-based auto complete in text editors. For routine tasks like generating {roxygen2} documentation scaffolding for functions, I use the {chores} package, which requires a bunch of pointing and clicking to use. Even though I use Positron, I purposely do not use either Positron Assistant or Databot. I have them disabled. So in the end, for pedagogical reasons, I don’t foresee me incorporating LLMs into this class. I’m pedagogically opposed to it. I’m facing all sorts of external pressure to do it, but I’m resisting. You’ve got to learn first.
December 9, 2025 at 8:17 PM
simonpcouch.com
A new release of odbc, a package that allows for connecting to various databases, is on #rstats CRAN! Just a few bug fixes in this release.

github.com/r-dbi/odbc/r...
A diagram containing four boxes with arrows linking each pointing left to right. The boxes read, in order, "R interface," "driver manager," "ODBC driver," and "DBMS." The left-most box, R interface, contains three smaller components, labeled "dbplyr," "DBI," and "odbc."
December 8, 2025 at 2:47 PM
simonpcouch.com
This edition of the @posit.co AI Newsletter highlights a few frontier model releases, a new R coding LLM evaluation dashboard, and a post from @wesmckinney.com on LLMs' difficulty with arithmetic!

posit.co/blog/2025-12...
2025-12-05 AI Newsletter - Posit
Several frontier model releases, the first viable AI-generated technical diagrams, and an evaluation suite for R code generation.
posit.co
December 5, 2025 at 3:50 PM
simonpcouch.com
I get the appeal of local LLMs—free, private, no big tech. That said, they're not yet ready to power coding agents like Positron Assistant or Databot.

www.simonpcouch.com/blog/2025-12...
Scatter plot revealing the cost-performance tradeoff across nine LLMs. Claude models occupy the upper-middle, with perfect accuracy and moderate cost. Gemini models tend to be cheaper but less performant. OpenAI models are more variable. The local (and budget Flash and Mini) models cluster in the lower-left, with minimal cost but zero performance. The models closest to the 'top left', being cheap but powerful, are Claude Haiku 4.5 and Gemini Pro 2.5.
December 4, 2025 at 2:27 PM
simonpcouch.com
It's Spotify Wrapped season🎁 Continuing the tradition of analyzing my music listening data with #rstats tidyverse, this time with Databot: www.simonpcouch.com/blog/2025-12...
A screenshot of the beginning of a conversation with databot, where I say 'Read blog/2025-12-03-wrapped-databot/Library.xml and turn it into a tibble' and the agent begins working.
December 3, 2025 at 9:57 PM
simonpcouch.com
A new release of vitals, a package for LLM evaluation in #rstats, is now on CRAN!🎄 This release includes all sorts of quality-of-life improvements; image support, precise latency measurement, better error messages, and more comprehensive logging.

github.com/tidyverse/vi...
December 2, 2025 at 2:17 PM
Reposted by Simon P. Couch
hadley.nz
I am tremendously grateful for open source, as it’s had such a huge impact on my career. That’s one of the reasons I’m so happy that Posit has committed to the @opensourcepledge.com. Read about our contributions to non Posit projects at posit.co/blog/posit-o....
Posit’s progress and renewed commitment to the Open Source Pledge - Posit
Posit continues its Open Source Pledge, investing heavily to support the open-source data science ecosystem.
posit.co
November 28, 2025 at 2:03 PM
Reposted by Simon P. Couch
rladiesparis.bsky.social
Everyone is welcome to join us at our next virtual meetup for a talk on using #GenerativeAI in R with Sharon Machlis
@smachlis.bsky.social — Tech Journalist and Data Professional.

RSVP 🔗 www.meetup.com/rladies-pari....

#RStats #RStatsFr #Rprogramming #RLadies #GenAI @rladies.org @posit.co
November 24, 2025 at 11:10 AM
Reposted by Simon P. Couch
ginareynolds.bsky.social
📣🚨 Very much looking forward to hearing from @frankiethull.bsky.social about the intersection of coded data viz and AI assistants! 👀🚀

Join us Tuesday Dec 2nd, at 3pm Eastern to hear about {ggpal2}, a {chores} extension. #ggplot2 #ai #chores

bit.ly/RxAI-join
November 21, 2025 at 9:54 PM
Reposted by Simon P. Couch
libbyheeren.bsky.social
The LLM privacy question comes up (rightly so) on every data community call I'm on. I've had sooo many questions myself. #databs #rstats #pydata
Good resource:
posit.co/blog/trust-l...

Also, have you seen side::kick in RStudio? simonpcouch.github.io/side/
Privacy and AI Assistants - Posit
Many of our LLM tools can execute code and access your files. What does this mean for privacy?
posit.co
November 21, 2025 at 7:40 PM
simonpcouch.com
The newest edition of the @posit.co AI Newsletter is out! Some notes on TOON, several open source updates, and a callout to two recent blog posts from @sara-altman.bsky.social and I.

posit.co/blog/2025-11...
2025-11-21 AI Newsletter - Posit
Several notable model releases from OpenAI, Gemini, and Moonshot, a new release of ellmer, and an LLM benchmark on plot interpretation.
posit.co
November 21, 2025 at 6:03 PM