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:...
I like the content.
But a small terminology note: I'm not sure it is a good practice to call it research pieces (and 80k hours is one of the norm-setting organization in EA)
A big portion of the pieces is podcasts, "recorded conversations between smart people". This is useful in many ways, but is it research?
In general, it seems to me ... "research" is high prestige in EA movement. Which creates an incentive to label things as research. So many important things are labelled research.
So it shouldn't surprise anyone there is e.g. shortage of operations people
Fair point. I thought about calling them articles... but they're definitely not all articles. Thought about calling them 'content releases' but that felt like corporate vagueness.
I should have gone with something nobody could dispute: 32 new(ish) universal resource locators. ;) - RW