Observed arrival · 2026-11-01
Haruka Kataoka’s AI livestream research
A graduate student’s portfolio of software projects and research, centered on an AI streamer designed to keep a livestream going.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- AI-streaming builders and speech researchers
- Worth noticing
- The stream’s decision layer is described as answering only yes or no; blinking and gaze are handled by a separate custom layer.
Field notes
The livestream description separates timing and flow from the writing, decision, speech, and face layers rather than treating the AI performer as one opaque model. Elsewhere, Kataoka lists a lab GPU setup with LLM, image, video, and speech generation, and a Slack notifier that checks for seminar openings hourly. These are project descriptions on the portfolio; the extract does not establish public access to the systems themselves.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
◉HumanPersonal, local, civic, or handmade
One card from the complete issue