Nice interview with Demis Hassabis (Deep Mind) about AGI timelines, security, LLM scaling, Gemini and what's coming youtube.com/watch?v=qTogNU…

Anthropic's Claud 3 System Card: www-cdn.anthropic.com/de8ba9b01c9ab7… It looks like it may be the most capable model on a few dimensions.

Apple is trying to catch up in the AI and LLM game. This week they published a research paper analysing the factors that make multi-model (image and text) LLMs work great. They deactivate different building blocks of the model architecture and estimate the impact on quality to

Do Norwegians put a higher weight on x-risks prevention? Their Oil Fund is going to be such a waste if we all blow up because of AGI, nukes, or bioweapons. Shouldn't a slice of the fund go towards x-risk research to make sure there will be future Norsks to enjoy it?

It was an awesome collaboration with the Azure Deployment Environments team! Excited to be their partner and look forward to our future projects!

Use ANY Terraform Provider from Pulumi! The entire Terraform / OpenTofu provider ecosystem is available to Pulumi users. And all the benefits of Pulumi are available to partners and developers building their own Terraform Providers, with no extra work! hubs.ly/Q02N50hd0

I'll be in Seattle Oct 20-24! Want to catch up over coffee? I'd love to hear from you - drop me a line!

What is state of the art for getting my personal/work context into Claude and ChatGPT? I looked at MCP but judging based on the quality of its GitHub server, it’s not really functional nor very actively improved.

Had to jump through the hoops but finally got Claude to search GitHub issues

The next bottleneck to scaling LLM training may be the desire to spend 100s of billions of dollars, there are no other obvious limits lesswrong.com/posts/NXTkEiaL…

A decision I made early in Pulumi was our best and worse decision: auto-naming. Every resource gets a random suffix appended to its physical name. This ensures • Guaranteed isolation from name collisions • Zero-downtime replacements Now thanks to Mikhail Shilkov, you have

And I closed an issue that I opened myself almost 7 years ago. My first ever issue in the Pulumi repo. So satisfying!

Excited to share what my team at Pulumi has been cooking! Our new Components feature lets you author infrastructure in one language and consume in any other. Simple API for users, complex machinery behind the scenes. Full blog post below!

Not yet sure if it’s a fool’s day joke or a next quarter deliverable

We launched a Pulumi MCP Server, the dev loop for your Pulumi projects has never been better pulumi.com/blog/mcp-serve…

I feel like I did something with AI github.com/pulumi/mcp-ser…

Dang. Windsurf with the Pulumi MCP server is nice! ✨

Compared Claude Code 1.x and 2.0 system prompts. Rigid "MUST/NEVER" rules became softer guidelines, some sections on code conventions removed, MultiEdit tool gone, SlashCommand tool added. Git safety got more specific. mikhail.io/2025/09/sonnet…

I poked at Claude Code’s web tools. Two paths: WebFetch (question → page answer) and WebSearch (discover links to fetch separately). Input/output schemas, implementation flow, and why it doesn’t do karaoke. mikhail.io/2025/10/claude…

New post: how Claude Code skills actually work: folder layout, the Skill tool, skills in the system prompt, and what’s injected at invocation time. Reverse-engineered from live sessions. mikhail.io/2025/10/claude…