Christopher Smith
@chriscarrollsmith.bsky.social
Software developer and technology consultant in Albany, NY
Title is admittedly intentionally provocative, but—here I lay out the functionalist case that a dynamic predictive text stream can be conscious and worthy of moral concern, and to some degree the current generation of models already are. open.substack.com/pub/knowledg...
Large Language Models are already conscious
... and this is the least conscious they'll ever be
open.substack.com
January 28, 2026 at 9:11 PM
Title is admittedly intentionally provocative, but—here I lay out the functionalist case that a dynamic predictive text stream can be conscious and worthy of moral concern, and to some degree the current generation of models already are. open.substack.com/pub/knowledg...
Safe to say you still need a GPU for this?
December 2, 2025 at 6:09 PM
Safe to say you still need a GPU for this?
At one point I was syndicating my Substack newsletter to my Github Pages site via RSS, and then they apparently blocked the Github Actions IP range from accessing RSS, so it stopped working.
November 20, 2025 at 11:53 AM
At one point I was syndicating my Substack newsletter to my Github Pages site via RSS, and then they apparently blocked the Github Actions IP range from accessing RSS, so it stopped working.
Used it for a bit and then stopped. Ran into too many issues like VSCode incompatibilities and nested renv environments installed inside each other. You can always use rix or devcontainers if you really need reproducibility.
November 6, 2025 at 4:06 AM
Used it for a bit and then stopped. Ran into too many issues like VSCode incompatibilities and nested renv environments installed inside each other. You can always use rix or devcontainers if you really need reproducibility.
They often take 4-5 days.
November 6, 2025 at 3:55 AM
They often take 4-5 days.
My first reaction when I read the one-paragraph summary this morning was "no", but then I researched it a bit and voted "yes".
November 4, 2025 at 10:24 PM
My first reaction when I read the one-paragraph summary this morning was "no", but then I researched it a bit and voted "yes".
I did some research this morning on our local town council and candidates and was really pleased with the non-partisan tone all the candidates struck. I'm not sure what the heck is going on in national politics, but here in my community, the candidates all seem to want to collaborate to solve stuff.
November 4, 2025 at 2:57 PM
I did some research this morning on our local town council and candidates and was really pleased with the non-partisan tone all the candidates struck. I'm not sure what the heck is going on in national politics, but here in my community, the candidates all seem to want to collaborate to solve stuff.
Reposted by Christopher Smith
This was a tough but necessary decision - I posted my own notes on this here, from the perspective of a current PSF board member simonwillison.net/2025/Oct/27/...
October 27, 2025 at 8:34 PM
This was a tough but necessary decision - I posted my own notes on this here, from the perspective of a current PSF board member simonwillison.net/2025/Oct/27/...
Many thanks to @christophscheuch.bsky.social and @tealemery.bsky.social, without whose support and vision this release would not have been possible.
October 25, 2025 at 7:53 PM
Many thanks to @christophscheuch.bsky.social and @tealemery.bsky.social, without whose support and vision this release would not have been possible.
New R package I developed as part of the `econdataverse` project for accessing macroeconomic data from the International Monetary Fund's IMF Data API was just published to CRAN. 🥳 Writeup here: open.substack.com/pub/modeling...
Announcing imfapi: User-Friendly Access to IMF Data in R
New IMF API, new IMF API R library — part of the econdataverse initiative
open.substack.com
October 25, 2025 at 7:51 PM
New R package I developed as part of the `econdataverse` project for accessing macroeconomic data from the International Monetary Fund's IMF Data API was just published to CRAN. 🥳 Writeup here: open.substack.com/pub/modeling...
They broke Twitter again; influx incoming
September 3, 2025 at 3:20 PM
They broke Twitter again; influx incoming
Bonds are fun to fight about, because yields go up when the economy is strong, but also when there's inflation or default risk.
Of course, you can tease apart what's driving them with a good multi-factor model, but it's a lot more fun to just vibe-Rorschach it and flame people on social media.
Of course, you can tease apart what's driving them with a good multi-factor model, but it's a lot more fun to just vibe-Rorschach it and flame people on social media.
August 20, 2025 at 1:33 PM
Bonds are fun to fight about, because yields go up when the economy is strong, but also when there's inflation or default risk.
Of course, you can tease apart what's driving them with a good multi-factor model, but it's a lot more fun to just vibe-Rorschach it and flame people on social media.
Of course, you can tease apart what's driving them with a good multi-factor model, but it's a lot more fun to just vibe-Rorschach it and flame people on social media.
Look at how Putin uses the handshake to establish dominance. He won't let go. Trump has to tap his hand to say, "we're done."
August 15, 2025 at 10:47 PM
Look at how Putin uses the handshake to establish dominance. He won't let go. Trump has to tap his hand to say, "we're done."
Answer: a few states (Minnesota, New Jersey) look a little better; a few others (South Carolina, New Mexico) look a bit worse.
But for the most part, the states permitting single-family housing and the states permitting multi-family are generally the same states.
But for the most part, the states permitting single-family housing and the states permitting multi-family are generally the same states.
August 6, 2025 at 1:40 PM
Answer: a few states (Minnesota, New Jersey) look a little better; a few others (South Carolina, New Mexico) look a bit worse.
But for the most part, the states permitting single-family housing and the states permitting multi-family are generally the same states.
But for the most part, the states permitting single-family housing and the states permitting multi-family are generally the same states.
Yesterday I ran across this map of housing permits by state. Seems to be single-family, so I wondered how much the map might be penalizing urban states that do more multifamily development.
To answer this question, I replicated the analysis with multifamily permits (next tweet).
To answer this question, I replicated the analysis with multifamily permits (next tweet).
August 6, 2025 at 1:40 PM
Yesterday I ran across this map of housing permits by state. Seems to be single-family, so I wondered how much the map might be penalizing urban states that do more multifamily development.
To answer this question, I replicated the analysis with multifamily permits (next tweet).
To answer this question, I replicated the analysis with multifamily permits (next tweet).
Thanks so much for the plug! I'm working on PDF parsing too, specifically with an aim to integrate it with a Zotero fork!
July 14, 2025 at 11:03 PM
Thanks so much for the plug! I'm working on PDF parsing too, specifically with an aim to integrate it with a Zotero fork!
A few days late, but here's my promised video coverage of the submissions: youtu.be/lQnBPqIpc2A
Chat with the budget bill in five CLI commands with `llm` — mini-hackathon recap, part 2
YouTube video by Christopher Smith
youtu.be
July 9, 2025 at 3:36 AM
A few days late, but here's my promised video coverage of the submissions: youtu.be/lQnBPqIpc2A
Which is to say, this was user error (well, really user laziness, because I knew we needed nullable fields but opted to skip it for purposes of the writeup), rather than model error. :)
July 7, 2025 at 10:35 AM
Which is to say, this was user error (well, really user laziness, because I knew we needed nullable fields but opted to skip it for purposes of the writeup), rather than model error. :)
Good catch. With the abbreviated schema syntax used to generate that output, you can't mark a variable in your schema as optional. That causes hallucinations by forcing the model to output a value when there shouldn't be one. You should instead use a fully specified JSON schema with nullable fields.
July 7, 2025 at 10:30 AM
Good catch. With the abbreviated schema syntax used to generate that output, you can't mark a variable in your schema as optional. That causes hallucinations by forcing the model to output a value when there shouldn't be one. You should instead use a fully specified JSON schema with nullable fields.
Gratitude to the participants, and many thanks to Simon Willison for creating and maintaining this amazing tool!
July 7, 2025 at 1:49 AM
Gratitude to the participants, and many thanks to Simon Willison for creating and maintaining this amazing tool!
*Winner*: The winning project, by Steve Senkus, scraped, chunked, and embedded the entire text of the "Big Beautiful Bill" for semantic search. Which, by the way, you can achieve with just five CLI commands!
July 7, 2025 at 1:49 AM
*Winner*: The winning project, by Steve Senkus, scraped, chunked, and embedded the entire text of the "Big Beautiful Bill" for semantic search. Which, by the way, you can achieve with just five CLI commands!
*Runner-up*: My own pitch was that you can use the help output from any command-line tool to create a prompt template for `llm`, then wrap the template with a shell script to create a purely natural language version of the tool. As proof of concept, I made `nl-repomix`:
July 7, 2025 at 1:49 AM
*Runner-up*: My own pitch was that you can use the help output from any command-line tool to create a prompt template for `llm`, then wrap the template with a shell script to create a purely natural language version of the tool. As proof of concept, I made `nl-repomix`:
*Runner-Up*: Evan Mullen scraped the HTML from Doctor of Credit, a website that aggregates bank deposit bonus offers, and then piped it to `llm` for conversion to a JSON array so he could compute which offers have the best ROI.
July 7, 2025 at 1:49 AM
*Runner-Up*: Evan Mullen scraped the HTML from Doctor of Credit, a website that aggregates bank deposit bonus offers, and then piped it to `llm` for conversion to a JSON array so he could compute which offers have the best ROI.
At my mini-hackathon on @simonwillison.net's `llm` tool, developers had just one hour to come up with a cool use case for `llm`. I wrote up what we came up with as a Github Gist. Highlights in the replies below! gist.github.com/chriscarroll...
llm-hackathon-submissions.md
GitHub Gist: instantly share code, notes, and snippets.
gist.github.com
July 7, 2025 at 1:49 AM
At my mini-hackathon on @simonwillison.net's `llm` tool, developers had just one hour to come up with a cool use case for `llm`. I wrote up what we came up with as a Github Gist. Highlights in the replies below! gist.github.com/chriscarroll...
Here's the demo presentation I gave to hackathon participants on some of the awesome things you can do with @simonwillison.net's `llm` command-line tool.
(This is a re-recording rather than the original because it was noisy in the venue.) youtu.be/UZ-9U1W0e4o
(This is a re-recording rather than the original because it was noisy in the venue.) youtu.be/UZ-9U1W0e4o
Become a command-line superhero with Simon Willison's `llm` tool — mini-hackathon recap, part 1
YouTube video by Christopher Smith
youtu.be
July 6, 2025 at 3:55 PM
Here's the demo presentation I gave to hackathon participants on some of the awesome things you can do with @simonwillison.net's `llm` command-line tool.
(This is a re-recording rather than the original because it was noisy in the venue.) youtu.be/UZ-9U1W0e4o
(This is a re-recording rather than the original because it was noisy in the venue.) youtu.be/UZ-9U1W0e4o