Hello, world ⎚

Personal website of Hazel Shanks ⎆

Socials ⏧

Decided to collapse a few of my identities, and admit that these are me:

Projects ⚙︎

🮶 The Grand Archive TCG card search site prereq.cards is mine.

🮶 fracturize is a 3D IFS / fractal-flame editor and renderer in development at https://github.com/HazelSh/fracturize

Documents ⎘

CV

My CV. Updated August 2026.
→ Hazel_Shanks_CV_(2026).pdf

Slides for: "Geopolitics of severe risk from frontier AI & New Zealand’s position"

Lightning talk for AISANZ 2026 conference. A brief discussion of frontier AI geopolitics through a NZ lens.
→ geopolitics_of_frontier_AI.pdf

Slides for: "What do you talk to when you talk to an LLM? The Persona-Selection Model, the Cyborgists, and Artificial Identity"

Topic was the identity of the interlocutor in LLM conversation, and how the Persona-Selection Model does and doesn't help understand the situation.
→ robopsychology-psm.pdf

Slides for: "The Doom Thesis — Why ‘If Anyone Builds It, Everyone Dies’: an ASI & x-risk primer."

Slides for the talk I gave at the Christchurch AI Safety & Ethics Meetup (Feb 16th 2026).
Topic was an intro to x-risk from ASI, covering Yudkowsky's book 'If Anyone Builds It, Everyone Dies' with regard to recent developments in AI.
→ doom_thesis_presentation_(extended).pdf

Slides for: "Can Claude Play Pokemon? Today's AI agent effectivenss, and the Time-Horizons view."

Slides for the talk I gave at the Christchurch AI Meetup (July 7th 2025).
Topic was AI agents and figuring out if/how they can actually get things done. I look at measuring that in the present moment (mid-2025), through the lens of the Claude Plays Pokemon public experiment.
→ claude-plays-presentation-(final).pdf

⛈︎