Dave Guarino
I wrote a new newsletter walking through the (very reasonable) idea maze of setting quality control metrics for a public benefits program — and how banal accounting details can lead to large structural forces. Click here or below to go give it a read!
There is so much excitement about things like GPT-4, and the Executive Summary Epistemology™ of broad, sweeping summaries of things people are saying about things other people are saying and which actually have very little grounding in tactile interactions with the technology are proliferating. In general, given just how new this wave of AI is, my perspective is that people should be spending more…
I wrote down some thoughts on accountability, feedback loops, Read these meandering thoughts over on Substack (I'll eventually cross-post the content here)
'a computer mainframe with duct tape on it in cyberpunk style' (Dalle) I really enjoyed two recent posts about engineering and LLMs: Mitchell Hashimoto's " Prompt Engineering vs. Blind Prompting " Apenwarr's " System design 2: what we hope we know " A theme that struck me as underlying both posts is nicely captured in the (rough) quote of the latter's engineering professor: Engineering isn't about…
I've decided to try out a newsletter using Substack . I plan to keep blogging here, but people have varying preferences in how they get stuff like this, so I'm considering the Substack to be: A way to do email distribution where I aggregate and share some things that may be here (or in toots/tweets) A bit of a test to find serendipity on that nascent network, which seems to be settling into a…
This post goes into the bucket of "draft" or "thinking out loud" I — like many others — am very excited by newly accessible large language models (LLMs) like ChatGPT and GPT-4 from OpenAI. I have been experimenting with these tools, and am rapidly developing a viewpoint that LLMs are about to change how we use software fairly dramatically, and fairly quickly. Recently I tried connecting OpenAI's…
I put a lot of intentional time into learning cooking during the pandemic. To be completely frank, I've never cooked much at all in my life, getting by mostly with frozen meals and cost-effective takeout like tacos (a blessing of living in California.) So this was a significant change for me. A lot of folks have talked about the parallels between cooking and programming. (Aside: I think it's not…
I spend a lot of time thinking about "practical knowledge." Or, a slight variation, how different forms of practice or participation in a system yield differential access to certain knowledge about a system. One category of knowledge on my mind lately is the understanding of edge cases when users of software are given some degree of flexibility: for example, allowing input on a front end that may…
No. There is no (documented) maximum limit for the number of issues a GitHub repository can have. Some repos with a large number of issues: Adguard Filters : ~113,000 Flutter : ~70,000 issues Kubernetes : ~41,000 issues Tensorflow : ~35,000 issues If you know of a documented limit or have evidence of an undocumented limit, please email me .
Recently, I shared a write-up of some work I worked on with a few wonderful colleagues (Anne and Justin) at the California Office of Digital Innovation, where I've done varied work for the past ~1.5 pandemic years or so. This work was fundamentally about identifying — in an acute and actionable way — barriers Californians were facing in accessing the Emergency Broadband Benefit (EBB), a pandemic…