alastair rushworth

weird one. flipped over to gpt-5 series of models and suddenly started using a lot more output tokens versus input - like millions more.

Data Scientist (with R) Blaze's rstats newsletter for R users and data scientists - covering new CRAN packages, tutorials and community updates

📊 This week in R

--> Positron IDE tour
--> vcr v2 enhancing usability
--> Quarto R package 1.5.0 with new workflow features

#rstats #datascience

blaze.email/Data-Scienti...

A simple thing I like - blaze newsletters now contain relevance filtered sections from company tech blogs. Example here is for the GenAI newsletter.

some improvements in look and feel on the newsletters this week.

I feel a bit silly learning so late that gmail doesn't always render CSS the way other clients do 🙃 anyways, if you use gmail, the emails should be looking much nicer now@

🤘 Download festival -- 2x weekend arena resale tickets for £240 each - sadly we can't make it 🥲 #download #DLXXII

secure.ticketmaster.co.uk/rs/360060CCF...

Sure it’s cool that AI can mimic any artist or writer. But what good is that if it consumes the business model that made the art / writing possible in the first place.

‘One day I overheard my boss saying: just put it in ChatGPT’: the workers who lost their jobs to AI From a radio host replaced by avatars to a comic artist whose drawings have been copied by Midjourney, how does it feel to be replaced by a bot?

Really good article. The saddest part to me is how much the AI companies have done to devalue human art and writing.

“The workers who lost their jobs to AI”https://www.theguardian.com/technology/2025/may/31/the-workers-who-lost-their-jobs-to-ai-chatgpt

Voter movements visualised to May 2025

Yes to both and especially the politeness point. For me it’s coming from the point of view of “I love your blog so much and want to read it as soon as you publish something”.

QQ for #rstats quarto blog owners - if your RSS feed was misconfigured / not configured, would you appreciate a rando like me politely pointed it out?

context - quarto by default doesn't enable rss feeds, and so there are a lot of amazing sites out there that can't be subscribed to.

and even after that time the desire to update.packages(ask=FALSE) still runs strong

a man with a mustache is sitting at a desk and saying it 's so ... it 's so beautiful . ALT: a man with a mustache is sitting at a desk and saying it 's so ... it 's so beautiful .

Just opened rstudio for the first time in maybe 2 years

Also… Quantum computing ⚛️

I’m really buzzed by the response to this newsletter and really happy with the content it surfaces.

So many cool blogs out there! A lot goes over my head (it’s been a while, I’m rusty ok 😅) but it’s scratching that itch

Experimenting with a new newsletter digest for the mathematically inclined. It contains a really good sweep of blogs - check it out (and let me know what you think)!

blaze.email/The-Mathemat...

reminder that you can search all 50k+ blog sites (posts too!) at blognerd.app. I forget about it from time to time even though I made it for myself 😅

But I think it's really cool to be able to find other people and their writing in really specific niches you are interested in.

Home Learn R, R tutorials, R resources, blog posts and the latest updates about the statistical programming R language.

Want to stay ahead in #RStats? My essentials: R Weekly (rweekly.org), #DataScience Community Newsletter (academicdatascience.org/resources/ne...), Blaze Tech (blaze.email), Carpentries Clippings (carpentries.org/about-us/new...), & rOpenSci (ropensci.org/news/). Any you'd recommend? Drop them here!

Some of the anti-AI stuff feels a bit like when people would say "don't use Wikipedia as a source." It's just like anything else, a piece of information that you weigh against multiple sources and your own understanding of its likely failure modes

📊 This week in R:
- R 4.5.0 released with faster package installations
- K-Means analysis comparing tech giants
- RSQLite with DBI database tutorial

#rstats #datascience

blaze.email/Data-Scienti...

Related youtu.be/pyXouxa0WnY?...

Fwiw: I use AI tools a lot and get a lot of benefit from them. But I also have concerns about the unscrupulous manner in which they are built, particularly with almost total disregard for IP and copyright laws across the world.

Very in vogue over here to make very general statements like this about the (non-) utility of LLMs & reductive statements about how they work.

More than ever we need intelligent and nuanced dissent and criticism, because the tech is moving almost too quickly to catch.

It might be nostalgia, but I’m sort of excited about the Digg reboot. reboot.digg.com

📊 3 R highlights:
1. R 4.5.0 released with new use() function for selective imports
2. New tempodisco package for temporal discounting analysis
3. ShinyConf 2025 coming soon

#rstats #datascience

blaze.email/Data-Scienti...

I'm obviously a little out of touch, as I didn't know about shinylive - that's interesting! yeah there's a bit of a learning curve to actually deploying a site. though I've found that with app-serving tools (I use digital ocean's) that hassle isn't as bad as what it was.

I think they'll always have their place, shiny is particularly great for making handsome scientific apps really fast with #rstats.

also, I really don't like the new 'vibes' coding meme - it suggests driving blind, which we should avoid. But I think a lot of the reason streamlit/shiny has been so popular is that it lowers the barrier to entry so much. I'd argue that code assistants do the same thing, just in a different way.

I always defaulted to streamlit / r shiny for making data apps because I'm a data person, and don't really know enough of any front end frameworks. I'm not sure what the future is for these tools now we have code assistants good enough to create a site using more general purpose frameworks.

📊 This week in R: data.table vs dplyr comparison, rOpenSci Champions Program 2025, and custom R package documentation with pkgdown #rstats #datascience

blaze.email/Data-Scienti...

Since Google took over Fitbit, the app has got worse and worse. This graphic in the app is an example. Completely meaningless.