Thoughts

Ideas, insights and perspectives on AI strategy and training.

Newsletters

A regular round-up of what we think you should be reading, AI tool updates worth knowing about, and practical insights to keep you current.

March 2026

Focus: From Experimentation to Execution

The agentic era arrives, AI changes the shape of work rather than reducing it, and confidence is outpacing readiness across most organisations. Plus tools from the frontline and our newly updated 2026 training.

Read more →

February 2026

Focus: Moving from AI Pilots to AI Performance

Three strategic AI trends shaping 2026, three tools we're using at Caversham House, and our newest training modules for leaders and teams moving from AI pilots to AI performance.

Read more →

January 2026

Focus: From Investment to Impact

AI investment is high but returns remain uneven, the people agenda takes centre stage, and the agentic future demands clarity of purpose. Plus three tools worth knowing about.

Read more →

Thought Leadership

Deeper perspectives on the strategic challenges of AI adoption.

March 2026

Philippa Cameron

Breaking Down the 2026 State of Data & AI Literacy Report

AI literacy is now a core workplace expectation, yet a significant skills gap persists. This article breaks down DataCamp's 2026 report and what it means for building genuine AI capability across your organisation.

Read more →

February 2026

Philippa Cameron

Leadership AI Training: Why Leaders Must Learn First

AI transformation fails when leaders treat it as something for IT to roll out. This article examines why leaders must invest in their own AI education first.

Read more →

January 2026

Chris Hornby

AI in 2026: solving the 6% problem

Nearly 90% of organisations use AI but only 6% capture meaningful returns - a gap that comes from workflow and organisation design, not from the technology.

Read more →

LinkedIn Posts

Short-form insights and commentary shared on LinkedIn.

Philippa Cameron

23 March 2026

Why peer sharing matters in self-paced learning

Peer sharing matters more in online learning because it is easy for people to complete a module, close their laptop, and move on without ever testing ideas in conversation.

Read more →

Philippa Cameron

18 March 2026

Start with the problem, and the tool becomes very helpful

Installing Claude CoWork and staring at the prompt box taught me a lesson I already knew: start with the problem, not the tool.

Read more →

Philippa Cameron

13 March 2026

When AI enters everyday workflows

When AI enters everyday workflows, the biggest change isn’t the technology but how work gets done, as tasks shift, steps disappear and new points emerge where human judgement matters more.

Read more →

Chris Hornby

11 March 2026

Where do we begin? Finding the bottleneck that pulls the whole chain faster

The most common question from clients isn't "what can AI do" but "where do we begin?" Map your team's repeated work, find the bottleneck step, and let an agentic tool turn it into a repeatable process.

Read more →

Chris Hornby

9 March 2026

Claude CoWork: conclusions and caveats | Post 6 of 6

CoWork is the first Agentic AI tool for the non-technical audience. Reflections on governance, knowledge silos, and what this all means for business users.

Read more →

Chris Hornby

9 March 2026

Claude CoWork: how task scheduling shifts the focus | Post 5 of 6

Scheduling work shifts the focus away from the AI tool entirely and onto the work itself - an interesting side effect of agentic AI.

Read more →

Chris Hornby

5 March 2026

Claude CoWork: When it connects to your world | Post 4 of 6

Connectors are the hands and eyes of Agentic AI - giving CoWork access to your emails, Slack, Monday.com, local files and even your web browser to do real work in your world.

Read more →

Philippa Cameron

5 March 2026

Building microlearning with the Claude Chrome extension

The Claude Chrome extension plugged directly into my learning platform was slow - but the real value wasn't speed. It was about delegating the grunt work and freeing up cognitive energy for what matters.

Read more →

Chris Hornby

4 March 2026

Claude CoWork: bringing your experience to the fore | Post 3 of 6

CoWork "skills" let you capture your professional experience as reusable instructions, making CoWork a specialist for any task you care about. And it even helps you build and test those skills itself.

Read more →

Chris Hornby

3 March 2026

Claude CoWork: Why it feels like a colleague | Post 2 of 6

Working together in a folder feels very different to working in a chat interface. When files appear on your machine, it becomes clear that real work got done - not just a conversation that went well.

Read more →

Philippa Cameron

3 March 2026

Reflections on the 2026 State of Data & AI Literacy Report

AI literacy has shifted from "nice to have" to foundational. The DataCamp report shows a 72% leadership consensus on its importance - yet a persistent readiness gap remains, and organisations with workforce-wide upskilling nearly double their AI ROI.

Read more →

Chris Hornby

28 February 2026

Agentic AI and Claude CoWork: what the fuss is about | Post 1 of 6

Agentic tools like Claude CoWork are probably going to change the way many of us work. In early 2026, Anthropic took the learnings from Claude Code and created CoWork: an agentic framework for thinking work in finance, marketing, sales and more.

Read more →

Chris Hornby

24 February 2026

Try Claude CoWork to understand Agentic AI

If you are curious about Agentic AI and how it differs from chatting to AI in your favourite tool, you need to try Claude CoWork. Buy a licence and experiment - it will be the best $20 you spend this month.

Read more →

Chris Hornby

12 February 2026

The "SaaSpocalypse": what the market got right and what it overreacted to

Anthropic's Cowork launch wiped $285 billion off SaaS stocks, but the sell-off conflated two very different threats: a genuine structural challenge for single-user workflow tools and mere budget pressure on enterprise coordination platforms.

Read more →

Philippa Cameron

11 February 2026

GenAI training and the search engine parallel

Updating our GenAI foundation course felt like the early 2000s all over again. The tools are different, but the capability challenge - prompting, iteration, critical thinking - is remarkably familiar.

Read more →

Chris Hornby

6 February 2026

Opus 4.6 and the frontier model race

Anthropic retakes the lead on frontier benchmarks with Opus 4.6. The competition is great for end users, but the elephant in the room is Microsoft - where most businesses have locked in their AI bet.

Read more →

Philippa Cameron

28 January 2026

Building AI tools with Cursor as a non-developer

Ten working tools built in a single day by a non-developer. What mattered most wasn't technical skill but clarity of intention, prompt iteration and thoughtful design.

Read more →

Chris Hornby

25 January 2026

Real ROI from AI: strategy plus capability

The companies seeing real ROI from AI do two things well: they are clear about what AI is for, and they give their people the ability to use it. One without the other does not compound.

Read more →

Philippa Cameron

20 January 2026

Why microlearning works in the AI age

Microlearning isn't just a trend - it reflects how learning actually works in modern organisations. Short, focused units help people build capability in the flow of work and apply knowledge immediately.

Read more →

Chris Hornby

19 January 2026

2026: the year of human-AI synergy

Klarna at scale and SMEs in day-to-day workflows are learning the same lesson: AI is strong where work is predictable, humans are strong where it requires judgment and empathy. The future is AI clearing the noise so people can do the work that matters.

Read more →

Chris Hornby

14 January 2026

Don't buy a 10 ton truck to deliver 4 tons of bricks

AI cost reduction is a sign of a maturing ecosystem. As models improve, the question shifts from "can AI do this work?" to "can a cheaper model do it at the same quality?"

Read more →

Chris Hornby

10 January 2026

LLMs are solving Erdős problems

LLMs are now adding to human knowledge in a real, quantifiable way. PhD-level talent is available at very low monthly cost - the challenge is how to turn that into ROI.

Read more →

Chris Hornby

5 January 2026

Plants grow in weird places - and so does AI

You can't predict where AI will take root. All you can do is be ready to capitalise when it does. The simplest silver bullet is investing in your people.

Read more →