I've been building out a sort of "clickwise"[1] for agents to talk to each other without resorting to English as a communication format.<p>My problem is approximately ~2000 rules in various levels of conditional nesting from a central context agent to the execution side of things.<p>This is a better worked example than my confused mess of financial rules.<p>The structure is a bit more fleshed out, but you are still sending across data which is hard to decipher for a human, but perfect for a machine to bootstrap and learn a set of nested rules about the task.<p>Finance is a bit less consistent than chess on the rules, but packing into binary "Telepathy" protocols is going to be the next big thing for multi-agent orchestration.<p>[1] - <a href="https://www.penny-arcade.com/story/automata/automata/automata-page-two" rel="nofollow">https://www.penny-arcade.com/story/automata/automata/automat...</a>
Why does the first frame contain 0, 1, 2, 3, 4, 7, 8 and 15 _twice_? It says something about leading zeroes, but visually that makes no difference. Why isn't this room used to put _all_ the values from 0 to 15 so that there's no implied question of why were those numbers chosen and no others and does that have meaning or not?
Am I wrong to immediately think of Contact?
wow!
I built Golden Board, a 512 KiB binary message designed to teach chess to an unknown recipient without shared language, culture, or prior knowledge of the game.<p>The project is inspired by a tradition of messages designed for an unknown intelligence such as Voyager Golden Record, Lincos and others.<p>But instead of sharing science or art, it teaches chess.<p>Why chess? Because a small set of rules opens up an enormous world of ideas and beauty. We can teach it from scratch and then let the recipient explore a rich legacy shaped by centuries of human ingenuity, sacrifices, and moments of brilliance.<p>The message begins by teaching the recipient to count, recognize bytes, and gradually decode and understand everything that follows.<p>I tested it extensively, using LLMs to simulate an unknown recipient and carefully checking each part of the message. Of course, LLMs are far from perfect stand-ins, but those experiments helped me identify ambiguities and improve the design.<p>The project is an attempt to solve a broader problem: its bootstrap is chess-agnostic and designed to decode other structured messages. Anyone is welcome to build on it and create messages that teach something entirely different.<p>Source code and docs: <a href="https://github.com/arlegotin/golden-board" rel="nofollow">https://github.com/arlegotin/golden-board</a>
[flagged]