
Building @aVentureVC (Techstars NYC, a company research platform using AI agents - https://xitter.cf/t.co/B2EnOfV0UY) & https://xitter.cf/t.co/iPpPJnFUAt. CFA, Angel/VC investor.
San Francisco, CA
Joined March 2009
- Tweets731
- Following4.9K
- Followers912
- Likes2.9K
Pinned Tweet
New year, new personal website!
With an added in-site terminal - williamcallahan.com – at San Francisco, CA
Ryzen SoC Linux users buckle up! @AMD Boosting AI/LLM Performance For @AMDRyzen @AMDRadeon iGPUs As Much As 18~23% With Linux 7.4
AMD PerfOpt benchmarks across Strix Halo (@FrameworkPuter Desktop), Strix Point, and Krackan Point hardware.
phoronix.com/review/amd-perf…
🚀 zg (zvec-grep) is now open source!
We built a local search tool for people and AI agents—
indexing and search run on-device by default, combining natural-language understanding with the speed and precision of rg.
🔒 Local-first: File processing, indexing, and search stay on-device by default, keeping your code and documentation private and secure.
⚡️ Ready in three steps: Install zg, index your workspace, and connect your agent. zg automatically configures supported agents and runs across macOS, Linux, and Windows.
🔎 Comprehensive search: Semantic search, BM25, hybrid retrieval, and rg work together across code, documentation, and data.
🤖 More efficient agents: On SWE-QA-Bench, tool calls and input tokens were both reduced by roughly half, while the judge score improved by 1.50 points.
GitHub: github.com/zvec-ai/zvec-grep
Blog: zvec.org/en/blog/2026-08-28-…
zg is still evolving. Try it, star the repo, and share your feedback—every suggestion helps make it better. ⭐️
This 👌
Yesterday I posted about Codex turning small tasks into big systems. Several people in the replies pointed to the same tool: 𝗣𝗼𝗻𝘆𝘁𝗮𝗶𝗹.
Besides adding rules to AGENTS.md, you can install this ready-made Skill. It’s like giving Codex / Claude Code a senior developer who’s too lazy to reinvent the wheel.
An example from the repo: you ask for a date picker, and the agent starts installing a library, writing a wrapper, and adding styles. Ponytail’s solution is one line:
If the browser’s built-in control does the job, you can skip all that extra work.
Before writing new code, Ponytail tells the agent to check these options in order:
→ Reuse an existing implementation in the codebase
→ Prefer the standard library and built-in platform features
→ Use an installed dependency if it solves the problem
→ If none of those are enough, write only what the current task needs
If the agent has already written a pile of code, you can also use 𝗽𝗼𝗻𝘆𝘁𝗮𝗶𝗹-𝗿𝗲𝘃𝗶𝗲𝘄 to check the current diff.
It flags unnecessary abstractions, duplicate implementations, and dependencies you don’t need. Each finding includes the file, line number, what to delete, and what to use instead. It returns a review checklist without changing the code.
Try it on the diff that keeps growing with every change.
𝗙𝗿𝗲𝗲 𝗮𝗻𝗱 𝗼𝗽𝗲𝗻 𝘀𝗼𝘂𝗿𝗰𝗲, with support for Codex, Claude Code, and Pi:
github.com/DietrichGebert/po…
From a talented team!
Introducing /skill-doctor v1
Skill doctor is a skill that improves your skills:
- Crawls your past CC, Codex, or Warp convos
- Scores them for efficiency and code quality
- Proposes diffs to your skills that you'd actually merge (STE-100 approved)
warp.dev/skill-doctor 🔖
Love this!
Git at Scale (by cursor) has been one of the most interesting blog posts i've read in a while. It came right when I was frustrated with Shopify's internal git system.
As an exercise, I've implemented it over the weekend as open source. It's a single rust binary that you can point at any S3 type object store. It uses WAL and CAS primitives and requires no other data store.
It also implements bundle-uri so large git repos (like our mono) are very fast to download as a chain of static bundles. Also comes with basic familiar UX.
github.com/tobi/walgit
Qwen 3.8 27b is a great open source model, certainly the best of its size yet.
williamcallahan.com/blog/qwe…
William Callahan retweeted
New York’s frontier vision lab General Intuition just raised a large Series A round of $320 million.
See their fundraising history, competitive analysis and more: aventure.vc/companies/genera…
General Intuition is also now hiring for several positions: jobs.ashbyhq.com/medal
William Callahan retweeted
Salesforce is acquiring @fin_ai (@ycombinator backed, S11 batch, formerly known as Intercom) for $3.6B.
Salesforce has a long history of acquisitions, and you can see them all with aVenture research: aventure.vc/companies/salesf…
techcrunch.com/2026/06/15/sa…
Anyone else just see like 50+ SF police and California state troopers come flying through San Francisco on I-280? Wonder where they were headed?
😢
The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.
The net effect of this order is that we must abruptly disable Fable 5 and Mythos 5 for all our customers to ensure compliance.
Access to all other Claude models is not affected.
We apologize for this disruption to our customers. We believe this is a misunderstanding and are working to restore access as soon as possible.
Read our full statement: anthropic.com/news/fable-myt…
Y Combinator’s co-founder match has a serious fake account problem these days. Fake LinkedIn profiles, etc.
The volume has exploded. The ratio of fake to real accounts is extremely high. It’s surprising to see.
A beautiful rust-based IDE TUI! Great work @TysonGeorge.
Replying to @dani_avila7
I've been loving Ghostty. I started the lazyide project because I've been wanting to spend all of my time in the terminal now. Check out lazyide: github.com/tgeorge06/lazyide
William Callahan retweeted
Opencode is one of the most elegant repos I’ve seen I recommend checking it out if you’re learning AI
How to make your AI product stickier to users
theverge.com/ai-artificial-i…
William Callahan retweeted
“Magic is just someone spending more time on something than anyone else might reasonably expect.”
– Teller
William Callahan retweeted
In my experience of life across personal, professional, and open source, I’m amazed at how many people kind of wait for things to happen to them instead of going out and making things happen for them. It’s a consistent difference between a certain level of success (not just financial).























