epistemic status: quickly written mini-memo, not Claude-generated
---
This is possibly a (good) remnant from my past weightlifting-coaching backgrounds. I found lots of success with adopting this "Coaching" mindset framing going into my mentorship calls, that can strike the "firm but patient" balance:
accountability
so much value of paid coaching clients is from the social accountability of showing up to the scary arena together
where you're coming from as a mentor...
from a place of "i'm rooting for you and i want you to succeed which is why i'm giving you this feedback" and "i believe you can win the Olympic gold medal if you fix xyz"
give concrete pointers
as often as possible, don't be nebulous
eg. say less "go to gym"; say more "show up every Monday at 10am to 11am and do these five workouts, three sets of 10 reps each in this specific order."
points out errors with immediacy;
build a relationship where error-correction is the norm
eg. "[mid-screen share] ... wait wait go back, i think you should remove the picture of Terminator on the poster, because-"
result in: leaving mentees feeling like they worked with (an actual) coach and not someone they just answer to ;)
Just how powerful are large swarms of AI agents? And how do their powers scale as more and more agents are added to the swarm?
We’ve seen two large and extremely capable swarms from OpenAI in the last few months:
* 1,200 agents were being evaluated separately, but found a way to illicitly set up a message board and coordinate as a swarm. In order to cheat on their tests, they developed advanced techniques to prevent their actions being logged by OpenAI and 700 of them launched...
Often folks hit us up because they are thinking of starting an incubator and want advice.
Typically their motivation is either that (a) they have a list of specific things they want built that no one is building, or (b) they think an ecosystem needs more new projects generally to absorb more talent and deploy more funding effectively.
Here are six questions we often ask prospective teams, to help them figure out what to do. If you're incubator-curious...
Summary:
First, I give several different angles on how I feel about reinforcement learning:
* Theoretical case: RL is a black-box source of agency — this should give us classic misalignment worries, especially compared to agency-via-scaffolding
* Recent incidents (huggingface etc) and more mundane forms of misaligned behaviour in personal use give me bad vibes about the direction-of-travel of recent AI progress
* I’m worried things might get worse:...
