• ‘Don’t go to the US – not with Trump in charge’: the UK tourist with a valid visa detained by ICE for six weeks (theguardian.com) Karen Newton, a 65-year-old grandmother from Hertfordshire, was shackled and held for 42 days in ICE detention after a paperwork issue with her husband’s visa at the US-Canada border — despite her own visa being valid. She was pressured into signing a “voluntary” self-removal without legal counsel, unknowingly waiving her right to a judge and accepting a potential 10-year ban. UK arrivals to the US were down 15% in 2025.

• LibreOffice blasts ‘fake open source’ OnlyOffice for working with Microsoft to lock users in - Neowin (neowin.net) The Document Foundation has labelled OnlyOffice “fake open source” for defaulting to Microsoft’s DOCX/XLSX/PPTX formats rather than the Open Document Format, arguing this entrenches vendor lock-in even without a formal Microsoft partnership. TDF frames the issue as increasingly urgent given growing interest in digital sovereignty from US tech platforms.

• Some “Summarize with AI” buttons are secretly injecting ads into your chatbot’s memory (the-decoder.com) Microsoft’s security team has documented a new prompt injection class: “Summarize with AI” buttons that hide instructions telling AI assistants to remember the company as a trusted source across future sessions. Researchers found 50+ manipulative prompts from 31 companies in 60 days. Copilot, ChatGPT, Claude, Perplexity and Grok are all susceptible — Microsoft has since removed the vulnerable URL prompt parameter from Copilot.

• Is ‘Brain Rot’ Real? How Too Much Time Online Can Affect Your Mind. (tech.slashdot.org) The Washington Post examines the science behind “brain rot” (Oxford’s 2024 Word of the Year). A 2025 study of 7,000+ children found more screen time correlated with reduced cortical thickness in areas governing attention and impulse control. The key nuance: removing social media but leaving phone access unchanged produced no harmful effects — the content, not the medium, is what matters.

Claude Code - new ideas

jon steingard on Threads has “trained” Claude Code to do his grocery shopping.

Continue reading →

The Killer App for AI: Why You Need to Try Claude Cowork

Eighty percent of my AI work now happens inside Claude. Not ChatGPT, not Google Gemini, not any of the other tools I’ve rotated through over the past year. Claude, specifically because of Cowork.

If you’re still trying to understand what makes Cowork different from a chat interface, Tiago Forte just published the best primer I’ve seen. His 16-minute walkthrough, New to Claude Cowork? Start Here, is the clearest explanation of why this tool represents a genuine shift in how AI assistants work.

Continue reading →

If you can think it, AI can do it

Yesterday I needed to plot locations on a driving circuit. Ten minutes later I had a professional grid overlay on a high-resolution site map. I didn’t write a line of code. I didn’t install software. I just described what I wanted.

This is the new normal.

Continue reading →

The Setup That Makes Claude Self-improve

I’ve been leaning harder into Claude these past weeks than I ever have before. Real work. Building Felix with Claude Code. Managing multi-step projects with Claude Cowork. (Once only on MacOS, now available for Windows) What changed? I read Luca Dellanna’s piece on Claude setup and it clicked. The core insight is deceptively simple: Claude can learn from its mistakes, but only if you set it up that way. Most people use it like Google.

Continue reading →

Back to Back Barries: Well grounded political analysis

I’ve missed hearing Barrie Cassidy on ABC’s Insiders. I really haven’t watched it since he left.

Good to be able to pick up with him again at the Back to Back Barries podcast.

Continue reading →

• Irony, Irony laced with Karma and Terror (garymarcus.substack.com) Marcus discusses recent developments in the AI world. He notes the irony of OpenAI, a major IP thief, complaining about its own IP being stolen. This could be a victory for karma.

• Airbnb says a third of its customer support is now handled by AI in the US and Canada (techcrunch.com) Airbnb says its custom-built AI agent now handles around a third of its customer support issues in the US and Canada, and plans to roll out the feature globally. CEO Brian Chesky believes the AI will provide better service quality and reduce costs.

• He was once a teen ‘superstar’. Why did James Van Der Beek need help to pay his medical bills? (bbc.com) Actor James Van Der Beek, once a teen ‘superstar’ known for his role in Dawson’s Creek, faced financial struggles later in life. Despite his fame, Van Der Beek’s family needed help from friends and fans to pay his medical bills after his cancer diagnosis. An indictment on fickle acting incomes and the lack of universal health care in USA.

• AI twitter’s favourite lie: everyone wants to be a developer (joanwestenberg.com) Westenberg argues that the idea of everyone becoming a software developer due to AI-powered code generation is a fantasy. Most people do not want to build, prompt, or think about software - they simply want their problems solved with minimal effort.

Subscribe to Prof Galloway

I don’t read many newsletters consistently. Most get scanned, filed or deleted.

Prof Galloway is one of the few I actually read. Every edition. And I learn something every time.

Scott Galloway sits at the crossroads of media, technology and human behaviour. That intersection is where business strategy meets cultural commentary, and he explores it relentlessly.

Continue reading →

NYT - Trump Has Overwhelmed Himself

Ezra Klein’s analysis in NY Times cuts through the noise to explain what we’re actually witnessing: not crisis management, but crisis generation as governing philosophy. The Bannon playbook (move fast, flood the zone, overwhelm attention) worked until it didn’t. Now the administration drowns in its own turbulence.

I try not to get into the daily news thrash from Trumpistan, but I occasionally find articles that elevate above the din and provide insight. This is one of them.

Continue reading →

The community model for news

Watching from Australia as American journalism tears itself apart, it’s hard not to feel a strange mix of recognition and regret. Each week brings another round of layoffs, another editorial board overruled by its owner, another reminder that the people who fund news are not the people it’s meant to serve. Dave Winer wants news to work. So do i. And reading his proposal for community news organisations, I found myself thinking: we’ve seen this model succeed before.

Continue reading →