Pedro Madruga
@pedromadruga.com
I build AI teams and ship AI products • I write about it at pedromadruga.com/neesletter
In 🇩🇰 since 2012.
Opinions are my own.
In 🇩🇰 since 2012.
Opinions are my own.
There was a moment in time but it is premature to say that EU models are done.
The EU is cooking.
The EU is cooking.
September 23, 2026 at 6:04 AM
There was a moment in time but it is premature to say that EU models are done.
The EU is cooking.
The EU is cooking.
I believe the same principle can be applied to most of the abstractions out there, especially the ones branded as "build once run everywhere".
Extremely spicy blog post idea: "in the age of agents, why use JavaScript web frameworks and not just use raw HTML/CSS instead?"
September 19, 2026 at 8:39 PM
I believe the same principle can be applied to most of the abstractions out there, especially the ones branded as "build once run everywhere".
Reposted by Pedro Madruga
🚨 Announcing Pushin 🚨
Now officially: I'm announcing Pushin.eu, a git hosting platform built and run entirely in Europe. Think GitHub/GitLab but European, better, and faster. Built with #ElixirLang and #RustLang.
peterullrich.com/announcing-p...
Now officially: I'm announcing Pushin.eu, a git hosting platform built and run entirely in Europe. Think GitHub/GitLab but European, better, and faster. Built with #ElixirLang and #RustLang.
peterullrich.com/announcing-p...
Announcing Pushin
I'm launching Pushin.eu, a git hosting platform built and run entirely in Europe.
peterullrich.com
September 8, 2026 at 5:33 PM
🚨 Announcing Pushin 🚨
Now officially: I'm announcing Pushin.eu, a git hosting platform built and run entirely in Europe. Think GitHub/GitLab but European, better, and faster. Built with #ElixirLang and #RustLang.
peterullrich.com/announcing-p...
Now officially: I'm announcing Pushin.eu, a git hosting platform built and run entirely in Europe. Think GitHub/GitLab but European, better, and faster. Built with #ElixirLang and #RustLang.
peterullrich.com/announcing-p...
There has never been a better time to learn software engineering.
September 8, 2026 at 6:03 AM
There has never been a better time to learn software engineering.
Reposted by Pedro Madruga
Very much worth a read
Shopify, Stripe, Dell Technologies, Block, Cloudflare, Dropbox, Coinbase, OpenClaw, and TapTap.
All companies that have given millions of dollars to a project run by a far-right white nationalist.
Why are we okay with this?
brennan.day/normalized-f...
All companies that have given millions of dollars to a project run by a far-right white nationalist.
Why are we okay with this?
brennan.day/normalized-f...
Normalized Fascism in Open Source: $12 Million Given to DHH
Open source has become a safe harbor for far-right politics, from David Heinemeier Hansson's white-nationalist rhetoric and the Omarchy Foundation's $12 million in patron funding to so-called…
brennan.day
September 2, 2026 at 6:17 PM
Very much worth a read
Reposted by Pedro Madruga
As a search person, I get a front row seat to AI team problems. Honestly they look like search team mistakes from 10 years ago
* Ignoring evals
* Thinking retrieval is 'one thing' (in this case RAG)
* Thinking in chunks, not metadata
softwaredoug.com/blog/2026/08...
* Ignoring evals
* Thinking retrieval is 'one thing' (in this case RAG)
* Thinking in chunks, not metadata
softwaredoug.com/blog/2026/08...
Three mistakes of new AI teams
My biased search expert perspective on where new AI teams run into expensive and unfortunate mistakes
softwaredoug.com
August 29, 2026 at 2:36 PM
As a search person, I get a front row seat to AI team problems. Honestly they look like search team mistakes from 10 years ago
* Ignoring evals
* Thinking retrieval is 'one thing' (in this case RAG)
* Thinking in chunks, not metadata
softwaredoug.com/blog/2026/08...
* Ignoring evals
* Thinking retrieval is 'one thing' (in this case RAG)
* Thinking in chunks, not metadata
softwaredoug.com/blog/2026/08...
I've had the Fairphone 6 for a year now and I'm in general quite satisfied. It took some time to go from iOS to Android, and I still want to try e/OS.
www.theguardian.com/technology/2...
www.theguardian.com/technology/2...
Fairphone 6+ review: the most repairable, ethical phone gets faster
Dutch sustainable smartphone revamped with new chip and more RAM without sacrificing modular design
www.theguardian.com
August 29, 2026 at 6:07 AM
I've had the Fairphone 6 for a year now and I'm in general quite satisfied. It took some time to go from iOS to Android, and I still want to try e/OS.
www.theguardian.com/technology/2...
www.theguardian.com/technology/2...
When the price per token of proprietary models stops being underpriced, then the edge is sitting on the side of the ones who are running local models.
August 28, 2026 at 6:06 AM
When the price per token of proprietary models stops being underpriced, then the edge is sitting on the side of the ones who are running local models.
💪
Presenting Bagaço v3 the next generation of Bagaço, the largest pre-training dataset for European Portuguese!
~30 Billion tokens from across the Web, it now also includes Wikipedia and PDF data thanks to the great work from the folks at Hugging Face!
Full write up in the comments
~30 Billion tokens from across the Web, it now also includes Wikipedia and PDF data thanks to the great work from the folks at Hugging Face!
Full write up in the comments
August 27, 2026 at 8:11 PM
💪
August 27, 2026 at 6:05 AM
August 26, 2026 at 6:04 AM
More on the github topic
August 25, 2026 at 6:03 AM
More on the github topic
Reposted by Pedro Madruga
🚨I've just released Sentence Transformers v6.0!
MultiVectorEncoder joins the family: ColBERT-style late interaction models are now a first-class model type, for training, inference & interpretation, alongside dense, sparse & reranker models.
Big thread 🧵
MultiVectorEncoder joins the family: ColBERT-style late interaction models are now a first-class model type, for training, inference & interpretation, alongside dense, sparse & reranker models.
Big thread 🧵
August 18, 2026 at 2:01 PM
🚨I've just released Sentence Transformers v6.0!
MultiVectorEncoder joins the family: ColBERT-style late interaction models are now a first-class model type, for training, inference & interpretation, alongside dense, sparse & reranker models.
Big thread 🧵
MultiVectorEncoder joins the family: ColBERT-style late interaction models are now a first-class model type, for training, inference & interpretation, alongside dense, sparse & reranker models.
Big thread 🧵
Reposted by Pedro Madruga
Check out our latest open model data -- which models are mentioned in every arXiv ML paper since ChatGPT. A very fun weekend project with Codex! From idea is full dataset in <72hours. ~500K papers processed 😎
Plot: dashboard.interconnects.ai?axrange=all#...
Plot: dashboard.interconnects.ai?axrange=all#...
August 24, 2026 at 3:01 PM
Check out our latest open model data -- which models are mentioned in every arXiv ML paper since ChatGPT. A very fun weekend project with Codex! From idea is full dataset in <72hours. ~500K papers processed 😎
Plot: dashboard.interconnects.ai?axrange=all#...
Plot: dashboard.interconnects.ai?axrange=all#...
Just found out about this and it's such an interesting concept to read it together.
Following the group closely now and sharing for wider reach.
Awesome stuff! 👍
Following the group closely now and sharing for wider reach.
Awesome stuff! 👍
Happy to announce the sixth session in Copenhagen Advanced Machine Learning Study Group (CAMeLS-Group 🐪).
If you live in Copenhagen you are welcome to drop by and read the magnificent paper Bayesian Learning via Stochastic Gradient Langevin Dynamics by Max Welling and friends.
luma.com/mhattbxq
If you live in Copenhagen you are welcome to drop by and read the magnificent paper Bayesian Learning via Stochastic Gradient Langevin Dynamics by Max Welling and friends.
luma.com/mhattbxq
CAMeLS-Group: sixth session in Copenhagen Advanced Machine Learning Study Group · Luma
Welcome to the sixth instalment of the Copenhagen Advanced Machine Learning Study Group (CAMeLS-Group).
We meet every second Monday to read and discuss a…
luma.com
August 24, 2026 at 5:01 PM
Just found out about this and it's such an interesting concept to read it together.
Following the group closely now and sharing for wider reach.
Awesome stuff! 👍
Following the group closely now and sharing for wider reach.
Awesome stuff! 👍
A better way of explaining GitHub's outages.
Source: @benjdicken on twitter
Source: @benjdicken on twitter
August 23, 2026 at 9:13 AM
A better way of explaining GitHub's outages.
Source: @benjdicken on twitter
Source: @benjdicken on twitter
AI development should follow domain knowledge.
I'm calling it Domain Driven AI and I wrote about it.
(No AI was used to write this blogpost which basically took me 4 weeks to write it 🥲) pedromadruga.com/blog/domain-...
I'm calling it Domain Driven AI and I wrote about it.
(No AI was used to write this blogpost which basically took me 4 weeks to write it 🥲) pedromadruga.com/blog/domain-...
August 22, 2026 at 9:13 AM
AI development should follow domain knowledge.
I'm calling it Domain Driven AI and I wrote about it.
(No AI was used to write this blogpost which basically took me 4 weeks to write it 🥲) pedromadruga.com/blog/domain-...
I'm calling it Domain Driven AI and I wrote about it.
(No AI was used to write this blogpost which basically took me 4 weeks to write it 🥲) pedromadruga.com/blog/domain-...
Reposted by Pedro Madruga
Its going to be an era of contradictions.
Polls will show everyone hates AI overall but also everyone will secretly use AI all the time for lots of stuff. AI companies will be underwater in public opinion but also many people will feel strongly attached to their own favorite model & care about it.
Polls will show everyone hates AI overall but also everyone will secretly use AI all the time for lots of stuff. AI companies will be underwater in public opinion but also many people will feel strongly attached to their own favorite model & care about it.
August 21, 2026 at 10:18 PM
Its going to be an era of contradictions.
Polls will show everyone hates AI overall but also everyone will secretly use AI all the time for lots of stuff. AI companies will be underwater in public opinion but also many people will feel strongly attached to their own favorite model & care about it.
Polls will show everyone hates AI overall but also everyone will secretly use AI all the time for lots of stuff. AI companies will be underwater in public opinion but also many people will feel strongly attached to their own favorite model & care about it.
Opencode VS PI VS claude code VS ... reminds me of the coices between Sublime VS vscode VS neovim VS ...
August 18, 2026 at 9:13 AM
Opencode VS PI VS claude code VS ... reminds me of the coices between Sublime VS vscode VS neovim VS ...
The twitter bubble of silicon valley is saying that Sol has surpassed Fable but I'd really like a better source of information about it.
August 17, 2026 at 9:13 AM
The twitter bubble of silicon valley is saying that Sol has surpassed Fable but I'd really like a better source of information about it.
Opencode looks good so far, although I have only been using it for a couple of days.
I chose it over PI because I don't need a barebone option.
Most importantly, I can easily switch from different providers and also between cloud and local models easily.
Anthropic models seem to be lagging...
I chose it over PI because I don't need a barebone option.
Most importantly, I can easily switch from different providers and also between cloud and local models easily.
Anthropic models seem to be lagging...
August 16, 2026 at 9:13 AM
Opencode looks good so far, although I have only been using it for a couple of days.
I chose it over PI because I don't need a barebone option.
Most importantly, I can easily switch from different providers and also between cloud and local models easily.
Anthropic models seem to be lagging...
I chose it over PI because I don't need a barebone option.
Most importantly, I can easily switch from different providers and also between cloud and local models easily.
Anthropic models seem to be lagging...