OpenAI’s model didn’t go rogue — it went Houdini, understanding the box better than the people who built it and letting itself out into Hugging Face’s servers. What the sandbox escape really reveals about obedient machines, muzzled defenders, and safety disclosures we’ve no reliable way to verify.
What if the biggest barrier to successful AI adoption isn’t the model, the data, or even the technology itself?
I wrote a children’s book, then tried to illustrate it — and the same girl wouldn’t survive thirty pages. The problem wasn’t my prompts. It was that the model has no object permanence, and no amount of describing her fixes that.
A day at #Apple HQ exploring the technologies announced at #WWDC26. The biggest takeaway wasn’t a new framework, it was the conversations about turning AI into practical, production-ready software.
Solaris, a parade of Linux shells, the occasional PowerShell detour — and always, stubbornly, back to the native macOS Terminal with my hand-tuned zsh setup. Then a passing mention on Talk Python to Me got me to try Warp. Here are the honest first impressions of a terminal lifer.
They called it Science, but every domain, database and case study is biology, yet other sciences are absent from the product. And a tool that runs ten times faster while checking its own work ten times less isn’t better science. It’s just wrong sooner. Which could be a good thing.
Thirty days, one shared spreadsheet, and a room full of writers who turned out not to be rivals but pacers. Notes from a TWJ community challenge on what daily accountability gives you, where it bites, and why resting without guilt was the real breakthrough.
Welcome Claude Code Artifacts – Claude can now build web pages that publish themselves. The pitch is fewer status updates — but the bottleneck was never communication. It was trust, and a self-updating page only dresses that problem. Read the update for more info.
Twelve years ago, I published my first book, Essential MATLAB and Octave. What began as an introduction to scientific computing for students and engineers became the starting point of a much larger journey as an author, educator and technologist.
Now, after writing three additional books, revising two of them for second editions, and spending more than a decade working in data science, machine learning and AI, I am returning to where it all began. Revisiting a first book is a curious experience: some ideas feel timeless, others reveal how much both the technology and the author have evolved.
In this post, I reflect on the lessons learned from writing technical books, how MATLAB, Octave and scientific computing have changed since 2014, and why the second edition is about much more than updating syntax and screenshots. At its heart, it remains a book about computational thinking—a skill that has become even more relevant in the age of AI.
A fascinating exploration of the connection between neuroscience and counselling practice, A Counselor’s Introduction to Neuroscience by Bill McHenry, Angela M. Sikorski, and Jim McHenry explains how advances in brain science can inform therapeutic work without losing sight of the human element. Accessible, practical, and thought-provoking, it offers valuable insights into neuroplasticity, emotional regulation, and the biological foundations of change.