Dave Guarino

A kernel I need to write up: the value of working in the open on the web (and best case being open source) has gone way, way up with the existence of frontier AI models today.

Nice little weekend thing: using Claude Code to fill out a bunch of forms sent to me that are non-fillable PDFs, otherwise requiring filling by hand. (And tell it to look at its output to iterate on placement!)

Step 1: ask Claude for stretch suggestions

Step 2: have nano banana generate an illustration of the stretch

I feel like people downplay that the major determinants of things happening in the world are:

- They are possible (eg technologically)
- They have a payout structure

Example: many forms of fraud meet 1, but not 2, and so generally don’t happen.

Building a real-time state update pipeline for 370,000 SNAP recipients with AI This is a crosspost of a write-up I did on the Propel blog about how we implemented a pipeline of monitoring the 50 state SNAP agencies’ websites for updates on the government shutdown, and piped it o...

New post: Building a real-time state update pipeline for 370,000 SNAP recipients with AI open.substack.com/pub/daveguar...

No actual policy changed has been announced, so people who get SNAP do not need to do anything until USDA or states announce something more.

I’ve seen a credible explanation here that:
1. AI curious folks use the paid tieMrs
2. The free tiers are much much worse, far behind the frontier
3. So people occupy different worlds

Google Gemini 3 Pro is out today. I ran through my standard "hard SNAP question" which I use to eval new models and it... absolutely excels.

(If you've got a hard public policy/benefits question you want to test, leave a comment and I'll run it through as I have time.)

My company, Propel, is doing a direct cash transfer to some high need SNAP families given November deposits are indefinitely delayed due to the shutdown.

Propel is contributing $1 million in company funds and looking to raise another $9m from partners.

www.propel.app/shutdown-fun...

claude_skills_audio_guide.md | Claude Comprehensive technical guide to Anthropic's Claude Skills system - architecture, implementation, best practices, and real-world applications for AI agents.

One awesome use of LLMs I've enjoyed is giving them a bunch of technical docs for a tool, and turning it into an audio-friendly text doc, which then I listen to with ElevenLabs Reader.

Here's an example, on Claude Skills claude.ai/public/artif...

“The AWS outage is affecting our daycare” is not something I expected to say.

Finished Dune. Worth continuing on to Messiah?

Yet another use of Claude Code I probably won’t go back from:

Working with an API I don’t know, just tell it to comprehensively read the docs and come up with some options for me

(Solved a very annoying subtle bad default for me last week)

We need to start talking about an AI benefit design And more specifically: how we design it to actually work

We need to start talking about what an AI jobs displacement benefit program would actually look like.

Yeah, I mean, they're all shape-of-system things at the end of the day. Just have to spend my waking moments translating for the less gonzo crowd I suppose.

An essay I'm simmering on — thinking about government's use of tech (& AI), disambiguating problems of:

- Capability
- Alignment
- Dysregulation (cc @dsquareddigest.bsky.social)

An important differential diagnosis for interventions! And one that I think is implicit (and/or wrong) too often.

AI Models Are Getting Dramatically Better at Complex Policy Questions: Evidence from SNAP Asset Limits Testing 45 AI models on SNAP eligibility questions reveals dramatic improvements. See how GPT, Claude, and Gemini evolved from harmful to helpful answers.

New research post! We tested ~45 AI models released over the past 18 months on a hard SNAP policy question (one with a meaningful potential harm profile.) We saw *dramatic* improvements in how they handled policy complexity.

www.propel.app/insights/how...

One meta thing with governance of gov use of AI I feel the traditional policy world is ignoring a bit is that the AI capabilities at the frontier actually can be a huge tool in doing oversight, monitoring, etc. scaling accountability

Kernels 02: Claude Compadre (Claude Code for non-coding things) Also: is curiosity all you need?

Kernels 02: Claude Compadre (Claude Code for non-coding things)
open.substack.com/pub/daveguar...

Having my newsletter emails say "Dave Guarino from Dave Guarino's Occasional Newsletter" has annoyed me long enough.

So I hereby announce my newsletter's new name:

"Plausible Legibility"

More progress on using Claude Code for my absolutely-personally-tailored daily todo list tool.

First bit was lots of keyboard shortcuts a la vim. Next I want a voice interface (via Apple Watch) so I've prototyped that with a text-based chat interface that will eventually be driven by voice.

I've been picking at an idea, that a corollary to "idea machines" in certain problem spaces is the creation of "agency machines"

In domains with high stuckedness, sometimes you need to inject a new actor with lots of ability (and incentive) to do things

I did it.

I waited out the need to actually learn the Javascript ecosystem.

Enjoying this AI coding future EVEN MORE upon this realization.

(lol I pay for Max...)

wrote my own static site LLM client in a few hours that does exactly what I want and nothing more. gpt-5, claude sonnet and opus, gemini 2.5 pro + web search + reasoning toggle. it rules. will make repo public soon

so good

I'm using Claude Code to add vim-style keyboard interactions to a personal todo app because THE COST OF PERSONALLY TAILORED SOFTWARE HAS BECOME $20 PER MONTH.

I’m beginning to suspect one of the highest societal value things to do right now given AI is to make the work displacement effects it will have (is having…) legible and indisputable.

I will go out on a limb and say that I do believe we should use new AI capabilities to increase the capacity of the state to deliver on the goals democratic governance processes have arrived at.

I'm increasingly convinced Claude Code is an early glimpse of what general purpose AI with agency feels like.

I'm experimenting with making it usable for non-programmers:

- Created a custom output style called "non-technical user"
- It explains what it's doing in lay terms