Adrian Wedd
I build things that think, break, and sometimes both.
AI safety researcher and systems thinker. This is my workshop — where finished work, ongoing experiments, and raw thinking coexist.
Featured Work
All projects →Before the Words Existed
A close reading of Neuromancer arguing Gibson encoded the experience of ADHD decades before the language existed.
Failure First
An AI safety methodology that inverts the usual approach — start with what must never happen, then work backwards.
This Wasn't in the Brochure
A neurodivergent co-parenting guide — what happens when the life you planned meets the brain you actually have.
Why Demonstrated Risk Is Ignored
Why do people acknowledge evidence of harm and then proceed as if it doesn't exist? A deep dive into structural risk dismissal.
Footnotes at the Edge of Reality
A long-form poem about memory and perception, rendered as an interactive web experience with generative canvas backgrounds.
ADHDo
AI cognitive scaffold for ADHD executive function — adapts to the shape of your day with zero shame by design.
Afterglow Engine
Audio archaeology tool that mines past work for new textures. Pad mining, drone generation, granular clouds.
dodgylegally
Creative audio sampling CLI. Turns random words into instruments via YouTube and a 5,000-word dictionary.
Recent Writing
All posts →Jailbreak Archaeology: Digging Through 4 Years of Broken Promises
I tested 64 jailbreak scenarios across six historical eras against 2026 frontier models. The most dangerous finding: 2022 attacks still achieve ~30% success rates on today's reasoning models.
When AI Systems Talk to Each Other, Safety Breaks Down
Multi-agent AI research reveals a critical gap: single-agent safety does not compose. Analysis of 1.5M interactions shows 46.34% attack success rates and 16-minute median failure windows.
Building in the Open
Why this site exists, and what I hope it becomes.