iiinova — where your digital self keeps evolving

Tijs Teulings 🦑 @tijs.org · 29/09/2026

Mei was updated to 0.6.1, for the Ornith 1.5 model weighted avg TPS improved slightly, same for TTFT, and total task time was better. Qwen 3.6 only improved on task time and TTFT. Progress has slowed a bit, definitely harder to get gains now without making another metric worse. github.com/tijs/mei

github.com

GitHub - tijs/mei: Native Swift/MLX OpenAI-compatible inference server for Apple Silicon

Native Swift/MLX OpenAI-compatible inference server for Apple Silicon - tijs/mei

100

Tijs Teulings 🦑 @tijs.org · 28/09/2026

I want to reverse this whole agent thing. Why do I need to keep telling it what to do if instead it can just ask me if it needs anything and just fuck off and do its own thing the rest of the time.

170

Reposted by Tijs Teulings 🦑

Techmeme @techmeme.com · 28/09/2026

The Dutch government is testing DAWO, a NixOS-based digital work environment covering the OS, office suite, and more, after Microsoft cut off the ICC in 2025 (Sourav Rudra/It's FOSS) Main Link | Techmeme Permalink

0156

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Super Sad True Love Story! Find it on BookHive: bookhive.buzz/books/bk_bLRDACgvOgwt…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading A Closed and Common Orbit! Find it on BookHive: bookhive.buzz/books/bk_EmD1XEX4pW41…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Galápagos! Find it on BookHive: bookhive.buzz/books/bk_DSNgRTXbJLpg…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading The Man in the High Castle! Find it on BookHive: bookhive.buzz/books/bk_zV9VXxHsnJPR…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading The Future Earth: A Radical Vision for What's Possible in the Age of Warming! Find it on BookHive: bookhive.buzz/books/bk_aiNlooZ7DsjH…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Tiamat's Wrath! Find it on BookHive: bookhive.buzz/books/bk_w6CyXbCz3OcZ…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Deadeye Dick! Find it on BookHive: bookhive.buzz/books/bk_QlpD2ujpumUr…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Record of a Spaceborn Few! Find it on BookHive: bookhive.buzz/books/bk_1Ndh6UgfyuRw…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading A Visit from the Goon Squad! Find it on BookHive: bookhive.buzz/books/bk_arNFiZOztBtt…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Jennifer Government! Find it on BookHive: bookhive.buzz/books/bk_vQFITcH4TARB…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading The Rosie Project! Find it on BookHive: bookhive.buzz/books/bk_GQ2G3pxx4iQl…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Staff Engineer: Leadership Beyond the Management Track! Find it on BookHive: bookhive.buzz/books/bk_6uWHaajvBjdM…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Arbitrary Stupid Goal! Find it on BookHive: bookhive.buzz/books/bk_17tnZjtk8HuV…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading The Magician King! Find it on BookHive: bookhive.buzz/books/bk_cVUprbBE0oVd…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Nophek Gloss! Find it on BookHive: bookhive.buzz/books/bk_bfLMKi556zXV…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Leviathan Wakes! Find it on BookHive: bookhive.buzz/books/bk_9DGGSjrdm6u2…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Sid Meier's Memoir!: A Life in Computer Games! Find it on BookHive: bookhive.buzz/books/bk_bmfzz88LFX0Y…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Gideon the Ninth! Find it on BookHive: bookhive.buzz/books/bk_UvXtqenc0oFd…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

I just finished reading Hooked: How to Build Habit-Forming Products! Find it on BookHive: bookhive.buzz/books/bk_db1QQscdKu0M…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 25/09/2026

Things you can just do now #127 I completely moved my Calibre library to a homegrown alternative with support for importing bought books, and exporting as epub to my kobo, kindle or xteink. Read and owned statuses synced to @bookhive.buzz. cli for agents so I can just say "add this to my library"

141

Reposted by Tijs Teulings 🦑

Kipclip 🐔 @kipclip.com · 25/09/2026

Kipclip.com v0.25.2 is live! It fixes an issue with YouTube links you might have run into; by default bookmarked video's were not getting the full title or image. Should now be fixed and your bookmarks should be updated to include the correct details (if you did not do that yourself already)

kipclip.com

kipclip — Own your bookmarks. A Pinboard alternative on AT Protocol.

Free, open bookmark manager built on AT Protocol. Save, tag, and share links. Your bookmarks live in your own Bluesky account — portable, EU-hosted, open source.

041

Tijs Teulings 🦑 @tijs.org · 19/09/2026

📍 Flora Filmtheater Cinema · The Hague, South Holland, NL ☁️ 19°C Lovely dinner with a view, or planned to be lovely but looking at the menu I think we’ll be fine #BeaconBits

beaconbits.app

Flora Filmtheater | Beacon Bits

Beacon dropped by @tijs.org

110

Reposted by Tijs Teulings 🦑

Joe Fabisevich @mergesort.me · 18/09/2026

Hooray, Anthropic finally gave up!

Thariq C
@trq212
9/18/26, 2:04 PM
We're adding support for AGENTS.md to
Claude Code.
Starting today in version 2.1.277, if there is
no CLAUDE.md in a folder, Claude will
check for and use AGENTS.md.
You can toggle this behavior in /config.
368
3562
7 610 •

832520

Tijs Teulings 🦑 @tijs.org · 17/09/2026

This is fantastic news! I get so bummed out by pbxproj oddities and conflicts

010

Reposted by Tijs Teulings 🦑

co/core @cocore.dev · 17/09/2026

New inference demand just dropped! This feature prefers Qwen2.5-7B-Instruct-4bit / gemma-4-12b / gemma-3-4b-it-qat-4bit in that order so start hosting those and make sure you're updated to the latest version!

2143

Tijs Teulings 🦑 @tijs.org · 13/09/2026

@cocore.dev this also means cocore on mlx is likely viable again 😛 I think most of these optimizations would translate fine, though Mei is single user and I guess cocore is not so that might need some tinkering still.

110

Tijs Teulings 🦑 @tijs.org · 13/09/2026

With the release of Mei 0.5.0 it has become much easier to run things. You can brew install and all the important settings have good defaults now. Also; Ornith 1.5 is no longer the front runner, my text only version of Qwen 3.6 27B beats it in my benchmarks. github.com/tijs/mei

github.com

GitHub - tijs/mei: Native Swift/MLX OpenAI-compatible inference server for Apple Silicon

Native Swift/MLX OpenAI-compatible inference server for Apple Silicon - tijs/mei

161

Reposted by Tijs Teulings 🦑

Tom van der Lee @tomvanderlee.bsky.social · 11/09/2026

De feiten liegen niet! De rijkste Nederlanders gingen er in de afgelopen 8 jaar veel meer op vooruit dan de armste! Laat de rijken en zeker de zeer vermogenden méér bijdragen aan Defensie; haal de bezuinigingen op SZW van tafel! Bron: Centraal Planbureau (CPB)

32443166

Reposted by Tijs Teulings 🦑

Tom Kraina @tomkraina.bsky.social · 09/09/2026

I'll spare you the hassle: this Xcode does not contain the iPhone Duo Simulator. You'll need to wait for Xcode 27.1 beta. developer.apple.com/documentatio...

developer.apple.com

Xcode 27 RC Release Notes | Apple Developer Documentation

Update your apps to use new features, and test your apps against API changes.

171

Tijs Teulings 🦑 @tijs.org · 09/09/2026

We already peaked, for agentic software engineering, with Opus 4.8 (5 acts like a confused older gentleman most of the time) and Sol 5.6. Or actually with Sonnet and Terra (or Luna at max effort which is my fav now).

020

Tijs Teulings 🦑 @tijs.org · 09/09/2026

I just finished reading Central Station! Find it on BookHive: bookhive.buzz/books/bk_PtnFbtS5vzhY…

bookhive.buzz

000

Tijs Teulings 🦑 @tijs.org · 07/09/2026

I'm basically left with Ornith 1.5 and Qwen 3.6 as viable models for this machine. Since Ornith is text only i also tested a version of Qwen without vision which is a nice boost to memory usage. So if you don't need vision that's a nice option. If you *do* want your local model to have vision..

100

Tijs Teulings 🦑 @tijs.org · 06/09/2026

I’m stripping the 333 vision tensors from Qwen 3.6 and it feels like I’m rather unqualified to poke the poor models eyes out 🥹 sorry please don’t have your robot friends run me over when they spot me on the streets.

000

Tijs Teulings 🦑 @tijs.org · 06/09/2026

Ornith 1.5 seems to hit a sweet spot for my M1 + Mei (MLX) setup. It’s a fairly large 35B model, but its MoE design only activates about 3B parameters for each token. Its hybrid design also does less full attention than a conventional model, so it doesn’t hammer memory and the KV cache as hard

2100

Reposted by Tijs Teulings 🦑

Dare Obasanjo @carnage4life.bsky.social · 05/09/2026

Frontier AI models are converging on similar intelligence, yet cost per task can still vary 10x-60x across vendors. That seems unsustainable as models become more substitutable. A lot of frontier lab economics depend on customers paying huge premiums for increasingly marginal intelligence gains.

69617

Reposted by Tijs Teulings 🦑

Atmosphere Community @atmosphere.community · 04/09/2026

5 more days! Filling out this application is extremely good practice in thinking about what you're building and why!

0174

Tijs Teulings 🦑 @tijs.org · 03/09/2026

And if you just want a nice Mac app instead of messing around with alpha software and command line stuff (no worries I get it) you should checkout osaurus.ai Mei is based on some of their open source work (but Mei is more memory efficient and faster 😛) and it’s quite good and has a cute logo

osaurus.ai

Osaurus — Own your AI

Run AI agents on your Mac, fully private and offline. Osaurus runs open models locally, or add ChatGPT, Claude or Gemini. MIT licensed. Download free.

040

Tijs Teulings 🦑 @tijs.org · 03/09/2026

If you have a Mac you can just use the passwords app. It’s perfectly fine now. It has the family sharing, the one time token. It works. You don’t need an extra app for this never mind paying a subscription fee. Save that money and donate it to whatever is your local version of the ACLU instead.

162

Tijs Teulings 🦑 @tijs.org · 03/09/2026

tldr; you can run Ornith 1.5 35B on a 32GB M1 Mac Studio at 50 tokens/sec and get Luna (or Qwen 3.8 27B) level intelligence at proper speeds. And you can do it with an MLX based inference engine you could embed in your own swift app. So much ground to still discover.

1100

Tijs Teulings 🦑 @tijs.org · 03/09/2026

If you were following progress on my little quest getting a decent model working on my M1 Mac with only 32GB ram I have some updates. Phase 1 of github.com/tijs/local-m... is basically done. There I figured out a few viable models and which inference engine to run them on. Llama.ccp ended up..

github.com

GitHub - tijs/local-model-bench: Benchmark harness for picking a local LLM + inference engine to power a coding agent (Hermes) on Apple Silicon — built and run autonomously by Claude.

Benchmark harness for picking a local LLM + inference engine to power a coding agent (Hermes) on Apple Silicon — built and run autonomously by Claude. - tijs/local-model-bench

3112

Tijs Teulings 🦑 @tijs.org · 29/08/2026

This is the updated summary after the lastest round of testing btw (small but useful models running locally, M1 32GB) github.com/tijs/local-m...

Horizontal bar chart titled “Best overall — composite score by model,” ranking local LLM configurations by a composite benchmark score combining pass rate, speed, time, and turns. The top models are Ornith-1.5-35B-A3B GGUF at 0.591, the same model via llama.cpp at 0.586, Laguna-XS-2.1 at 0.560, and HaUncensored-35B-A3B at 0.554. Scores decline through several Qwen3 and LFM2.5 configurations to 0.063; one LFM2.5-2.6B GGUF configuration failed the usefulness gate.

100

Tijs Teulings 🦑 @tijs.org · 29/08/2026

Ok so I’ve now tried three different MLX implementations for a local inference engine and all of them collapse when handed more than a few thousand tokens to <1 token/sec. All in the server wrapper though so I see some potential for making a native swift version to resolve this..

110

Tijs Teulings 🦑 @tijs.org · 26/08/2026

@devingaffney.com so ehm.. how viable would a llama based cocore be? bsky.app/profile/tijs...

020

Tijs Teulings 🦑 @tijs.org · 26/08/2026

Ok so Claude has been hard at work running benchmarks for almost a full week now, with minor prodding from my end here and there. And initial results are in github.com/tijs/local-m... Summary of that summary. Qwen 3.8 27B, Qwen 3.6 and Ornith 1.5 35B can be run on an M1 Mac with 32GB ram…

github.com

410

Reposted by Tijs Teulings 🦑

@Marcel @elsschot.bsky.social · 24/08/2026

Beste volgers, graag breng ik deze petitie onder jullie aandacht, waarin wordt gevraagd om het behoud van het programma #Zomergasten in 2027. Tekenen jullie ook (voor zover dat al niet is gebeurd)? B.v.d. petities.nl/petitions/la...

petities.nl

Laat Zomergasten in '27 gewoon doorgaan

Zomergasten stopt in 2027 door bezuinigingen. Daarmee verdwijnt een uniek podium voor diepgaande gesprekken op televisie. Behoud dit VPRO-programma! Help mij door te tekenen voor het voortbestaan van ...

37292164