flâneur

Using tweets gathered from https://til.simonwillison.net/twitter/collecting-replies

gist.github.com · 6,705 words · saved by 1 readers

Group these into a set of definitions that illustrate the various different shapes of definition. For each of your definitions include quotes of the tweets that support that definition. Answer in markdown Absolutely! Here are the definitions of "agent" (in the context of AI and LLMs) grouped by common themes, along with supporting quotes from the tweets you provided: Definition: Agents are AI systems that can independently work towards achieving a defined goal. They can plan, make decisions, and adapt without constant human intervention. Supporting Quotes: Definition: Agents are often characterized by their ability to use tools or external systems (APIs, functions, etc.) to accomplish tasks. They leverage LLMs to decide when and how to use these tools. Supporting Quotes: Definition: Agents are applications where the sequence of actions or control flow is not predetermined but is dynamically decided by an LLM based on the context and goal. Supporting Quotes: Definition: Agents are often

[ { "datetime": "2024-10-07T14:10:49.000Z", "username": "theprincessxena", "tweet": "A large language model hooked up to tools that lets it do useful things", "href": "https://twitter.com/theprincessxena/status/1843292901960524047", "likes": 9, "impressions": 869, "retweets": 0 }, { "datetime": "2024-10-07T14:29:17.000Z", "username": "maciejgryka", "tweet": "A state machine affecting some external system, which uses an LLM to make some of its decisions.", "href": "https://twitter.com/maciejgryka/status/1843297550176240110", "likes": 1, "impressions": 75, "retweets": 0 }, { "datetime":…

saved by

related reading