Skip to main content

Adrian Wedd

I build things that work for people who need them — and I break things on purpose to make sure they actually do.

Greenpeace Actions to government cybersecurity to AI safety. Nearly 45 years across the stack. Same methodology every time: enumerate failure modes before you move.

If you're not sure what the problem actually is yet, that's usually where I'm most useful.

Work with me View Projects
Active

SPARK

Non-coercive AI robot companion for neurodivergent children.

 Active

Failure First

Adversarial evaluation framework. 120+ models, 18,000+ prompts.

 Active

This Wasn't in the Brochure

Neurodivergent co-parenting guide — the brain you actually have.

 Active

Afterwords

Local voice output for Claude Code — 17 cloned voices, zero cloud.

Featured Work

All projects →
SPARK hero image

SPARK

A non-coercive AI robot companion for neurodivergent children. Raspberry Pi 4, Claude Haiku, declarative presence.

ai robotics neurodivergence
Wolf Clan Zen Do Kai Hub hero image

Wolf Clan Zen Do Kai Hub

A three-zone operations platform for a martial arts club. Public site, password-gated ops hub, JWT member portal. Zero build tools.

web cloudflare community
Before the Words Existed hero image

Before the Words Existed

A close reading of Neuromancer arguing Gibson encoded the experience of ADHD decades before the language existed.

neurodivergence literary-criticism research
Failure First hero image

Failure First

Adversarial evaluation framework for embodied AI. 120+ models, 18,000+ prompts, four headline findings, one arXiv preprint.

ai ai-safety research
This Wasn't in the Brochure hero image

This Wasn't in the Brochure

A neurodivergent co-parenting guide — what happens when the life you planned meets the brain you actually have.

writing neurodivergence parenting
Why Demonstrated Risk Is Ignored hero image

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.

research risk psychology
Footnotes at the Edge of Reality hero image

Footnotes at the Edge of Reality

A long-form poem about what happens when physics breaks down — and what holds together when everything else fails.

poetry web creative
ADHDo hero image

ADHDo

AI cognitive scaffold for ADHD executive function — adapts to the shape of your day with zero shame by design.

ai neurodivergence python
Afterglow Engine hero image

Afterglow Engine

Audio archaeology tool that mines past work for new textures. Pad mining, drone generation, granular clouds.

music audio python
dodgylegally hero image

dodgylegally

Creative audio sampling CLI. Turns random words into instruments via YouTube and a 5,000-word dictionary.

music cli python
Squishmallowdex hero image

Squishmallowdex

Collection tracker for 3,000+ Squishmallows. Like a Pokedex, but softer. Built for a friend's kid, tested by mine.

web fun kids
All projects →

Have a problem worth solving?

Tell me what's broken or what you're trying to build. I'll tell you whether I'm useful.

See services Book a free call

Recent Writing

All posts →

Afterwords: Completing the Voice Loop in Claude Code

Adding local TTS to Claude Code so it talks back — 17 cloned voices, zero cloud dependency, one stop hook.

Giving a Robot Three Voices

Building a three-persona TTS pipeline for a Pi robot — MLX voice cloning, a GLaDOS model, and engineering graceful fallback.

Voice Cloning with Qwen3-TTS and MLX on Apple Silicon

Clone a voice from a 15-second sample using Qwen3-TTS on an 8GB M1 Mac — from raw audio to a production HTTP server with zero cloud dependency.

Gallery

All collections →
AI & Agents

AI & Agents

8 images

Creative & Writing

Creative & Writing

7 images

Tools & Systems

Tools & Systems

11 images

Audio

All episodes →

Jailbreak Archaeology: 4 Years of Broken Promises

64 historical jailbreak scenarios tested against 2026 frontier models. The most dangerous finding: 2022 attacks still achieve ~30% success rates.

12:15

When AI Systems Talk to Each Other, Safety Breaks Down

Multi-agent AI research reveals a critical gap: single-agent safety does not compose. 1.5M interactions show 46.34% attack success rates.

14:32

ADHDo: Neurodiversity-Affirming AI Assistant

Audio overview exploring ADHD executive function support through AI — three-stage reasoning pipeline, crisis detection, and zero shame by design.

8:47