early stage product: 馃 collect tinder, start fire before freeze
later stage product: 馃敟 keep fire alive, grow fire, defend fire
very different
h/t @christineluc
any watermarking ai researchers who can explain this?
what if the llm should
- repeat something verbatim (important in a compaction prompt)
- there is just one correct order of tokens for a somewhat long chain (a certain sequence of control signals)
- asking for a diff of 2
stripe can offer a 'sign in with openrouter'
(stripe tokens?)
let users bring their own token supply to a service (whether b2b, b2c etc)
also s/users/agents