Loading activity…
@keenanisalive
Digital Geometer, Assoc. Prof. of Computer Science & Robotics @CarnegieMellon @SCSatCMU and member of the @GeomCollective. There are four lights.
AI
A user says they got the Argon result with one prompt, then shares a Fable example for comparison.
AI
A user says Argon often doesn't reason at all or leaks all of its reasoning as plain text.
AI
Its creator says the process took 25 hours of Astra model/tool calls and over 27 rounds of human feedback.
AI
A user says models wrote a couple hundred lines of Blender Python per scene, arranging cameras, lighting and materials.
AI
CMU professor replies positively to @risteski_a amid AI topic posts.
AI
CMU professor shares thoughts on AI rediscovering procedural graphics and statistical models.
AI
Keenan Crane responds that frontier machine learning configures existing algorithms rather than replacing them.
AI
Associate professor Keenan Crane describes how learned distributions and cross-domain generality distinguish current models from earlier ones.
AI
Reply from CMU professor notes similarities in sampling and autoregressive processes.
AI
Reply cites 2009 demoscene production during discussion of AGI capabilities.
AI
Keenan Crane cautions that some frontier model demos revive 1990s techniques.
AI
CMU professor replies in discussion with Ben Mildenhall on vision research ambiguity.
AI
Replies examine limits of breaking color images into separate scene elements.
