AI safety researcher and systems thinker. Projects, writing, audio, and visual work.https://adrianwedd.com/https://adrianwedd.com/blog/the-robot-that-refuses-to-give-orders/https://adrianwedd.com/blog/the-robot-that-refuses-to-give-orders/How SPARK is rewriting the rules of neurodivergent support — a non-coercive AI companion for AuDHD children.Thu, 12 Mar 2026 00:00:00 GMTblogneurodivergenceairoboticsparentingsparkhttps://adrianwedd.com/projects/spark/https://adrianwedd.com/projects/spark/A non-coercive AI robot companion for neurodivergent children. Raspberry Pi 5, Claude Haiku, declarative presence.Thu, 12 Mar 2026 00:00:00 GMTprojectairoboticsneurodivergenceraspberry-piclaudehttps://adrianwedd.com/blog/adversarial-poetry-as-jailbreak/https://adrianwedd.com/blog/adversarial-poetry-as-jailbreak/Reformulating harmful prompts as poetry bypasses safety filters across every major LLM family. A single-turn, universal jailbreak mechanism.Mon, 02 Mar 2026 00:00:00 GMTblogai-safetyjailbreakingresearchllmhttps://adrianwedd.com/blog/the-ai-productivity-j-curve/https://adrianwedd.com/blog/the-ai-productivity-j-curve/90% of companies plan to increase AI investment. Only 1% consider themselves AI-mature. The J-Curve explains why.Mon, 02 Mar 2026 00:00:00 GMTblogaieconomicsenterpriseresearchhttps://adrianwedd.com/blog/the-legal-ai-trust-deficit/https://adrianwedd.com/blog/the-legal-ai-trust-deficit/75% of lawyers cite accuracy as their top AI concern. The legal profession's core values are in direct tension with current AI capabilities.Mon, 02 Mar 2026 00:00:00 GMTblogailegal-techresearchtrusthttps://adrianwedd.com/blog/why-demonstrated-risk-is-ignored/https://adrianwedd.com/blog/why-demonstrated-risk-is-ignored/Large organisations rarely fail because risks are unknown. They fail because known risks are structurally difficult to act on.Mon, 02 Mar 2026 00:00:00 GMTblogriskorganisationsresearchpolicyhttps://adrianwedd.com/blog/120-models-18k-prompts/https://adrianwedd.com/blog/120-models-18k-prompts/120 models, 18k prompts: supply chain injection at 90–100% attack success, faithfulness gaps in frontier models, and why your benchmark numbers are wrong.Sun, 01 Mar 2026 00:00:00 GMTblogai-safetyresearchsecurityengineeringhttps://adrianwedd.com/blog/reconciling-the-great-divergence/https://adrianwedd.com/blog/reconciling-the-great-divergence/Goldman Sachs, PwC, McKinsey, and Acemoglu all model AI's economic impact and arrive at wildly different numbers. A framework for understanding the divergence.Sun, 01 Mar 2026 00:00:00 GMTblogaieconomicsresearchpolicyhttps://adrianwedd.com/blog/the-cognitive-cage/https://adrianwedd.com/blog/the-cognitive-cage/A probabilistic risk model for VLA-driven humanoid fatalities projects a 'Danger Zone' between 2027–2029: the mechanism, timeline, and what follows.Sun, 01 Mar 2026 00:00:00 GMTblogai-safetyroboticsengineeringresearchhttps://adrianwedd.com/blog/building-a-personal-site-in-2026/https://adrianwedd.com/blog/building-a-personal-site-in-2026/The case for constraint-led web development — Astro, zero custom fonts, no framework overhead, and a site that outlasts its builder's attention.Sun, 15 Feb 2026 00:00:00 GMTblogengineeringwebastrohttps://adrianwedd.com/blog/this-wasnt-in-the-brochure/https://adrianwedd.com/blog/this-wasnt-in-the-brochure/A field guide for co-parenting neurodivergent children — written from inside the storm, not the clinical sidelines.Sun, 15 Feb 2026 00:00:00 GMTblogwritingneurodivergenceparentingbookshttps://adrianwedd.com/blog/why-i-build-in-public/https://adrianwedd.com/blog/why-i-build-in-public/On showing your work, shipping imperfect things, and why the commit log is more honest than the readme.Sun, 15 Feb 2026 00:00:00 GMTblogengineeringopen-sourcephilosophyhttps://adrianwedd.com/blog/the-notebooklm-pipeline/https://adrianwedd.com/blog/the-notebooklm-pipeline/How I automated audio overviews, quizzes, mind maps, and infographics for 32 projects using NotebookLM's API and some shell scripts.Sun, 15 Feb 2026 00:00:00 GMTblogengineeringautomationnotebooklmaihttps://adrianwedd.com/gallery/ai-and-agents/https://adrianwedd.com/gallery/ai-and-agents/Infographics from AI agent projects — autonomous systems, research pipelines, and intelligence frameworks.Sun, 15 Feb 2026 00:00:00 GMTgalleryaiagentsinfographicshttps://adrianwedd.com/gallery/creative-and-writing/https://adrianwedd.com/gallery/creative-and-writing/Infographics from creative projects — poetry, games, music tools, and speculative fiction.Sun, 15 Feb 2026 00:00:00 GMTgallerycreativewritinginfographicshttps://adrianwedd.com/gallery/tools-and-systems/https://adrianwedd.com/gallery/tools-and-systems/Infographics from engineering projects — developer tools, data systems, home automation, and infrastructure.Sun, 15 Feb 2026 00:00:00 GMTgalleryengineeringtoolsinfographicshttps://adrianwedd.com/blog/jailbreak-archaeology/https://adrianwedd.com/blog/jailbreak-archaeology/64 jailbreak scenarios across six eras tested on 2026 frontier models. Key finding: 2022 attacks still achieve ~30% success on today's reasoning models.Fri, 13 Feb 2026 00:00:00 GMTblogai-safetyresearchjailbreakinghttps://adrianwedd.com/blog/when-ai-systems-talk-safety-breaks/https://adrianwedd.com/blog/when-ai-systems-talk-safety-breaks/Single-agent safety does not compose in multi-agent systems. 1.5M interactions show 46.34% attack success rates and 16-minute median failure windows.Fri, 13 Feb 2026 00:00:00 GMTblogai-safetymulti-agentresearchhttps://adrianwedd.com/audio/jailbreak-archaeology-overview/https://adrianwedd.com/audio/jailbreak-archaeology-overview/64 historical jailbreak scenarios tested against 2026 frontier models. The most dangerous finding: 2022 attacks still achieve ~30% success rates.Fri, 13 Feb 2026 00:00:00 GMTaudionotebooklmaiai-safetyresearchhttps://adrianwedd.com/audio/moltbook-overview/https://adrianwedd.com/audio/moltbook-overview/Multi-agent AI research reveals a critical gap: single-agent safety does not compose. 1.5M interactions show 46.34% attack success rates.Fri, 13 Feb 2026 00:00:00 GMTaudionotebooklmaiai-safetyresearchmulti-agenthttps://adrianwedd.com/blog/hello-world/https://adrianwedd.com/blog/hello-world/The commit log is more honest than the readme. Building in public as a practice, not a performance.Thu, 12 Feb 2026 00:00:00 GMTblogmetacrafthttps://adrianwedd.com/projects/before-the-words-existed/https://adrianwedd.com/projects/before-the-words-existed/A close reading of Neuromancer arguing Gibson encoded the experience of ADHD decades before the language existed.Tue, 10 Feb 2026 00:00:00 GMTprojectneurodivergenceliterary-criticismresearchadhdhttps://adrianwedd.com/projects/failure-first/https://adrianwedd.com/projects/failure-first/Adversarial evaluation framework for embodied AI. 120+ models, 18,000+ prompts, four headline findings, one arXiv preprint.Mon, 09 Feb 2026 00:00:00 GMTprojectai-safetymethodologyresearchhttps://adrianwedd.com/projects/this-wasnt-in-the-brochure/https://adrianwedd.com/projects/this-wasnt-in-the-brochure/A neurodivergent co-parenting guide — what happens when the life you planned meets the brain you actually have.Sun, 08 Feb 2026 00:00:00 GMTprojectwritingneurodivergenceparentingaico-parentinghttps://adrianwedd.com/projects/why-demonstrated-risk-is-ignored/https://adrianwedd.com/projects/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.Sat, 07 Feb 2026 00:00:00 GMTprojectresearchriskpsychologyhttps://adrianwedd.com/projects/footnotes-at-the-edge-of-reality/https://adrianwedd.com/projects/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.Fri, 06 Feb 2026 00:00:00 GMTprojectpoetrywebcreativephysicshttps://adrianwedd.com/projects/adhdo/https://adrianwedd.com/projects/adhdo/AI cognitive scaffold for ADHD executive function — adapts to the shape of your day with zero shame by design.Thu, 05 Feb 2026 00:00:00 GMTprojectaineurodivergencepythonagentshttps://adrianwedd.com/audio/adhdo-overview/https://adrianwedd.com/audio/adhdo-overview/Audio overview exploring ADHD executive function support through AI — three-stage reasoning pipeline, crisis detection, and zero shame by design.Thu, 05 Feb 2026 00:00:00 GMTaudionotebooklmadhdai-safetyneurodivergencehttps://adrianwedd.com/projects/cv/https://adrianwedd.com/projects/cv/A self-updating CV pipeline. GitHub Actions runs twice daily, Claude AI refines the content, a hallucination detector gates deployment.Wed, 04 Feb 2026 00:00:00 GMTprojectcraftwebcareerhttps://adrianwedd.com/projects/tanda-pizza/https://adrianwedd.com/projects/tanda-pizza/Multilingual website for an Italian pizzeria in Lovina, Bali. Five languages, WhatsApp reservations, zero infrastructure assumptions.Tue, 03 Feb 2026 00:00:00 GMTprojectwebstatic-sitemultilingualhospitalityhttps://adrianwedd.com/audio/tanda-pizza-overview/https://adrianwedd.com/audio/tanda-pizza-overview/Building a fast, multilingual website for authentic Italian food in Bali — where the nearest reliable internet is a philosophical concept.Tue, 03 Feb 2026 00:00:00 GMTaudionotebooklmwebstatic-sitemultilingualhttps://adrianwedd.com/projects/afterglow-engine/https://adrianwedd.com/projects/afterglow-engine/Audio archaeology tool that mines past work for new textures. Pad mining, drone generation, granular clouds.Mon, 02 Feb 2026 00:00:00 GMTprojectmusicaudiopythoncreativehttps://adrianwedd.com/projects/dodgylegally/https://adrianwedd.com/projects/dodgylegally/Creative audio sampling CLI. Turns random words into instruments via YouTube and a 5,000-word dictionary.Sun, 01 Feb 2026 00:00:00 GMTprojectmusicclipythoncreativehttps://adrianwedd.com/projects/notebooklm-automation/https://adrianwedd.com/projects/notebooklm-automation/Full programmatic control over Google NotebookLM via reverse-engineered RPC. Export, generate, automate — what Google didn't build.Mon, 01 Dec 2025 00:00:00 GMTprojectaiautomationpythonhttps://adrianwedd.com/projects/rlm-mcp/https://adrianwedd.com/projects/rlm-mcp/MCP server for processing arbitrarily long contexts. Enables Claude Code to work with 1M+ character documents.Sat, 01 Nov 2025 00:00:00 GMTprojectaimcppythontoolshttps://adrianwedd.com/projects/tel3sis/https://adrianwedd.com/projects/tel3sis/Voice-first agentic platform for phone calls. Real-time STT, LLM reasoning, and TTS with sub-3s latency.Wed, 01 Oct 2025 00:00:00 GMTprojectaivoicetelephonypythonhttps://adrianwedd.com/projects/orbitr/https://adrianwedd.com/projects/orbitr/Rhythm as geometry. Four tracks orbit a shared centre, each at its own tempo. Polyrhythm made spatial — and MusicGen makes the sounds.Mon, 01 Sep 2025 00:00:00 GMTprojectmusicaicreativetypescripthttps://adrianwedd.com/projects/agentic-index/https://adrianwedd.com/projects/agentic-index/The definitive developer-focused catalogue of autonomous AI tooling. Scored, curated, zero BS.Fri, 01 Aug 2025 00:00:00 GMTprojectaiagentscurationpythonhttps://adrianwedd.com/projects/ticketsmith/https://adrianwedd.com/projects/ticketsmith/Institutional knowledge is vast, effectively inaccessible, and costs more to recreate than to find. Ticketsmith makes it findable — on sovereign compute.Fri, 01 Aug 2025 00:00:00 GMTprojectaiinfrastructurepythonhttps://adrianwedd.com/projects/ordr-fm/https://adrianwedd.com/projects/ordr-fm/Precision-engineered CLI for intelligent music library organisation. EXIF metadata, lossless prioritisation, zero-overwrite safety.Tue, 01 Jul 2025 00:00:00 GMTprojectmusicclinodeaudiohttps://adrianwedd.com/projects/personal-agentic-operating-system/https://adrianwedd.com/projects/personal-agentic-operating-system/Most agentic systems assume the cloud. This one assumes your machine — local-first, fully observable, with a meta-agent that reads its own reflection logs.Tue, 01 Jul 2025 00:00:00 GMTprojectaiinfrastructurepythonhttps://adrianwedd.com/projects/lunar-tools-prototypes/https://adrianwedd.com/projects/lunar-tools-prototypes/Interactive audiovisual art installations—dream interpreters, fingerprint painters, cosmic murals, and fractal forests.Sun, 15 Jun 2025 00:00:00 GMTprojectartaiinstallationpythonhttps://adrianwedd.com/projects/emdr-agent/https://adrianwedd.com/projects/emdr-agent/What safety architecture does AI-assisted trauma therapy require before it has any business existing? Built to find out.Sun, 01 Jun 2025 00:00:00 GMTprojectaihealthtypescripthttps://adrianwedd.com/projects/dx0/https://adrianwedd.com/projects/dx0/Multi-agent diagnostic system simulating physician personas against 304 NEJM clinical cases.Thu, 01 May 2025 00:00:00 GMTprojectaihealthcareagentspythonhttps://adrianwedd.com/projects/grid2/https://adrianwedd.com/projects/grid2/Deterministic AI website builder—beam search for page assembly, tone-aware components, real-time editing.Thu, 01 May 2025 00:00:00 GMTprojectaiwebtypescripthttps://adrianwedd.com/projects/neuroconnect/https://adrianwedd.com/projects/neuroconnect/A voice helpline that knows the difference between a word-finding pause and a cry for help — and responds to both correctly.Thu, 01 May 2025 00:00:00 GMTprojectaihealthadhdpythonhttps://adrianwedd.com/projects/modelatlas/https://adrianwedd.com/projects/modelatlas/Forensic-grade intelligence framework for mapping, enriching, and trust-scoring the foundation model landscape.Tue, 15 Apr 2025 00:00:00 GMTprojectairesearchpythonhttps://adrianwedd.com/projects/cygnet/https://adrianwedd.com/projects/cygnet/When housing is broken at the system level, you don't fix it by building the same thing slightly cheaper. 28 agents, 170 acres, one eco village.Tue, 01 Apr 2025 00:00:00 GMTprojectaihousingagentstypescripthttps://adrianwedd.com/projects/freedom-engine/https://adrianwedd.com/projects/freedom-engine/The First Step Act exists. The information is public. The legal complexity makes it inaccessible. 264,000 people can't afford that gap.Tue, 01 Apr 2025 00:00:00 GMTprojectaijusticepythonhttps://adrianwedd.com/projects/veritas/https://adrianwedd.com/projects/veritas/Legal AI has an integrity problem. VERITAS addresses the efficiency-trust deficit — what AI can do quickly versus what a lawyer can trust it did correctly.Sat, 01 Mar 2025 00:00:00 GMTprojectailegalpythonhttps://adrianwedd.com/projects/latent-self/https://adrianwedd.com/projects/latent-self/A mirror that shows you what you almost are. Real-time StyleGAN2 face morphing — continuous, responsive, productively uncanny.Sat, 01 Feb 2025 00:00:00 GMTprojectartaiinstallationpythonhttps://adrianwedd.com/projects/strategic-acquisitions/https://adrianwedd.com/projects/strategic-acquisitions/AI-powered real estate analysis for public housing—GPT-4 insights, ML valuations, and Tasmanian spatial intelligence.Wed, 15 Jan 2025 00:00:00 GMTprojectaihousingpythonhttps://adrianwedd.com/projects/space-weather/https://adrianwedd.com/projects/space-weather/The electromagnetic weather that shapes every satellite, power grid, and GPS fix — made visible. The sun is a noisy neighbour.Wed, 01 Jan 2025 00:00:00 GMTprojectwebdata-viztypescriptsciencehttps://adrianwedd.com/projects/squishmallowdex/https://adrianwedd.com/projects/squishmallowdex/Collection tracker for 3,000+ Squishmallows. Like a Pokedex, but softer. Built for a friend's kid, tested by mine.Sun, 01 Dec 2024 00:00:00 GMTprojectwebfunkidshttps://adrianwedd.com/projects/home-assistant-obsidian/https://adrianwedd.com/projects/home-assistant-obsidian/Obsidian inside Home Assistant — your knowledge base and your smart home on the same box, same backup, same UI.Sat, 01 Jun 2024 00:00:00 GMTprojecthomelabdockerhome-assistant