It’s time to take your ideas seriously.
You can now build apps directly in Fuser:
↳ a design tool for your favorite effects
↳ a game that gets the group chat to meet IRL
↳ a photo diary of textures to use later
Make it for you and your people.
Go live in minutes.
New Anthropic Fellows research: the Assistant Axis.
When you’re talking to a language model, you’re talking to a character the model is playing: the “Assistant.” Who exactly is this Assistant? And what happens when this persona wears off?
this started with a striking PC1 falling out of persona space
my main insights from the past few months:
⊹ “distance from the Assistant” is the main axis of persona variation across these models e.g. the most relevant thing seems to be “how Assistant-like is this persona”
⊹ this axis already exists in base models and steering with it makes them speak from the POV of helpful archetypes like therapists, coaches, and consultants
⊹ not all personas far from the Assistant are bad! the risk comes from departing the more predictable territory of post-trained behaviour
still have a lot of questions about what to anthropomorphize, what to treat as fundamentally alien…
@vgr the webpage hosting the interview that you did w/ troy conrad therrien for the architecture effects exhibition seems to have been removed. any chance you have a copy of the PDF? It's a favorite of mine and would love to revisit it. thanks!
I am looking forward to collect new types of expensive rock. With expensive rock I want to be hit myself on the head, with it. I am making money moves, to be getting lot of money to with which purchase rare expensive rock to hit my head over and over with the rock. Oooohhhuhhhahh
true Radical Decentralized ubi That I Think Could Potentially Be Achieved Thru Crypto And Gaming But I Haven’t Ironed That Idea Out Enough Yet To Explain has never been attempted.