Verifying AI in your global security stack
As security systems become more automated, an AI error in one tool can disrupt the rest down the chain.
Trusted information for people in harm's way.
I combine technology and journalism to build mission-critical products in global security and disaster response.
Co-founder and chief product officer at Factal.
Where I focus
Products and companies that help people act on trusted information in critical situations.
More in projects →Journalistic methods for verifying what's true when speed matters and mistakes have consequences.
More about verification →Practical applications, emerging risks and the human judgment needed to use AI well.
More about AI →Speaking
I speak about practical AI for global security and risk teams.
Drawing on my work as a journalist and product builder, I bring real-life examples and practical approaches to keynotes, panels and training workshops.
Writing
AI and the anxiety of endless ideas
AI makes divergence nearly free while convergence remains hard. For most of my life, ideas felt scarce enough to save.
The disruptive product has to leave the building
Google had ChatGPT a year before launch and was too nervous to ship it. That's the Innovator's Dilemma in a single tweet.
Learning from AI that doesn't work
Many try AI, get poor results and give up. The better mindset is to hold yourself accountable for the output.
An untold part of the Breaking News story
Writing up the projects page, I paused on Breaking News — there's a part of that story I never published.