Pinned
Your users have suddenly acquired the ability to speak code into existence. What if your app gave them a safe place to run it? jeremymorrell.dev/blog/extensi...
Jeremy Morrell
@jeremymorrell.dev
Principal engineer @ Cloudflare working on Observability Things Location: Squamish, BC 🇨🇦 Website: https://jeremymorrell.dev/
3.3K followers743 following6K posts
Pinned
Your users have suddenly acquired the ability to speak code into existence. What if your app gave them a safe place to run it? jeremymorrell.dev/blog/extensi...
Jeremy Morrell reposted
Notes on Kimi K3.1 - this one is a big deal, the first time an open weights (well, soon-to-be open weights) model has competed this credibly with the current frontier simonwillison.net/2026/Oct/23/...
Dashboards are going away. They were junk drawers anyway. I'm still confused as to what an o11y product looks like today, but this is getting closer burcs.dev/notes/interf...
Jeremy Morrell reposted
whoah.. uh.. guys?? Kimi K3.1 Pro is eerily close to Opus 5.5 and also open weights and also 3T www.kimi.ai/blog/kimi-k3...
Jeremy Morrell reposted
Formal methods becoming orders of magnitude easier to do is a huge deal for software security.
Co-signed. This is my employer so apply salt to taste, but if your PHP service was behind Cloudflare the first rules protecting you went up ~9 hours before disclosure blog.cloudflare.com/php-vulnerab...
- shrink your surface and offload as much of that surface as you can to competent vendors (Cloudflare, AWS, etc) - I don't think you need abliterated models for this; GLM 5.3 is happy to do a lot of poking at stuff, just ask @hailey.at - reluctantly: give the frontier labs some money too
My existential crisis about “what even *is* an o11y product in 2026” deepens
As of September 2026, over 60% of our weekly active users were connecting via MCP -- running queries, creating dashboards, responding to alerts. That's a pretty wild number, given that our MCP server officially released just a year earlier.
Jeremy Morrell reposted
Hello from #AGNTCon! I'm presenting about what we learned building the @honeycomb.io MCP server. If you'd like to follow along, follow this thread! Otherwise, feel free to mute it. Let's get it started!
Jeremy Morrell reposted
Jeremy Morrell reposted
This is probably my favorite blog post on the CF blog this year. I've never gone from "wait, it's a what?" to "omg, I need to use this for so many things" so quickly I guess CF is a model lab now? 🤔
Introducing Clairic, from Cloudflare. A System One extraction modeI, capable of extracting, computing, estimating, and forecasting data from unstructured inputs. https://cfl.re/3UcTaMP
Jeremy Morrell reposted
Introducing Clairic, from Cloudflare. A System One extraction modeI, capable of extracting, computing, estimating, and forecasting data from unstructured inputs. https://cfl.re/3UcTaMP
Jeremy Morrell reposted
Jeremy Morrell reposted
Jeremy Morrell reposted
Yesterday, the CSSWG accepted my proposal for mesh gradients and diffusion curve gradients! Want to make gradients like this natively in CSS? Well, we've started the work on it. It's coming! Here's the Explainer: github.com/w3c/csswg-dr...
Jeremy Morrell reposted
This matches my recent experiences. I can ask for things, and the output works, but it's all wrong in the details. Not in a "doesn't fit my preferences" way, but in a "makes the system vastly more complicated and leaks that out to the user" way.
I gave Opus 5.5, considered the best model these days, a medium-sized task. Even after a full adversarial review and simplify loop to convergence, it produced code that worked but was garbage, with limited use of newtypes and representable invalid states everywhere
Microsoft did their own TPUs? blogs.microsoft.com/blog/2026/10...
This honestly seems like a sea change? classcad.ch I think it’s difficult to overstate how much impact more accessible CAD systems could have…
In retrospect it should have been obvious that “as long as you have a reference implementation to check against you can get LLMs to rebuild basically anything” would mean new CAD kernels 👀 github.com/NVIDIA-Omniv...
Another point against the “if AI is so great, where’s all the new software??” argument
In retrospect it should have been obvious that “as long as you have a reference implementation to check against you can get LLMs to rebuild basically anything” would mean new CAD kernels 👀 github.com/NVIDIA-Omniv...
In retrospect it should have been obvious that “as long as you have a reference implementation to check against you can get LLMs to rebuild basically anything” would mean new CAD kernels 👀 github.com/NVIDIA-Omniv...