Notes from the work.
Practical, plain-language posts on working with AI, making video, writing B2B content and building the systems behind it. A new post every Tuesday.
Latest
More to read.
Creator lessonsWhat 261 million views taught me about contentLessons from building a children's animation channel on my own.WritingWriting social posts that make one point, fastThe rules I follow for B2B posts that busy people actually read.Working with AIBuild an AI knowledge base your whole team can useThe difference between an AI that guesses and one that knows your business.AI videoFrom picture book to 3D: bringing your own art to life with AIWhy I start AI animation from characters I drew years ago.B2B contentThe one-person marketing team: systems behind 100+ posts a yearHow I keep a full content calendar running on my own.AI videoPlan your AI video one scene at a timeShort AI clips reward the people who plan. Here is how I break a video down.ProposalsHow I cut RFP setup from two days to 30 minutesWhat changed when I stopped treating every proposal as a fresh start.B2B contentMaking cybersecurity funny without making it sillyHow characters and humour carry a serious message for a B2B IT firm.
Also on LinkedIn
Shorter thoughts.
I also post on LinkedIn about using AI inside real companies.
LinkedInGovernance is what makes enterprise AI trustworthy
The true differentiator is not building models that talk; it is building the governance that makes them safe enough to trust.Read on LinkedIn →LinkedInWhy AI rollouts add admin fatigue
Sustainable AI enablement is never about asking knowledge workers to chat with models.Read on LinkedIn →LinkedInFrom navigation menus to conversation
The future of enterprise digital experience is not prettier page layouts; it is conversational systems that understand user intent.Read on LinkedIn →LinkedInWhy AI projects stall after the demo
Clever prompts produce articulate generalities. Reliability comes from a structured knowledge base.Read on LinkedIn →LinkedInFix the knowledge before you fine-tune
An LLM is not a company knowledge base. It is an inference engine, and it struggles when it is fed scattered, uncurated documents.Read on LinkedIn →
Coming up
Next on the blog.
- Making cybersecurity funny without making it silly
- How I cut RFP setup from two days to 30 minutes
- Plan your AI video one scene at a time
- The one-person marketing team: systems behind 100+ posts a year
- From picture book to 3D: bringing your own art to life with AI
- Build an AI knowledge base your whole team can use
- Writing social posts that make one point, fast
- What 261 million views taught me about content