--- schema: 1 type: video title: >- If we want them to do Knowledge Work, design them as Knowledge Agents — Benjamin Clavié, Mixedbread description: >- An agent searching a contract for "30 days" has no way to know whether it has found a deadline, a grace period, or a retention rule, or wandered into a medication schedule entirely. Benjamin Clavié uses that to explain why coding agents turned out to be the easy case. Code carries durable cues: identifiers, file paths, method definitions, things you can literally grep for. It also comes with narrow instructions, because nobody asks a coding agent to invent a paradigm, they ask it to close a ticket. Knowledge work outside code is the opposite on every axis. The meaning is implicit, the same string means different things in different domains, and the search has to start from an intent the agent must reconstruct on its own. His argument is that the industry generalized from the one domain that least resembles the rest. The good news is that humans solved this already, repeatedly. Clavié traces a tool loop running from the first catalogue of the Library of Alexandria through bibliographies to search engines, and an organization loop running from the single gifted polymath through monasteries and universities to the modern specialist firm. They are really one loop: better tools create new roles, new roles generate knowledge, new knowledge demands better tools. His benchmark numbers make the practical version concrete. A badly tuned retrieval tool lands near 60 percent accuracy, which nobody would trust enough to keep using. A well tuned one reaches the same ceiling as its rivals while making 20 percent fewer tool calls, and adding a layer of searcher agents that return short memos closes roughly 40 percent of the remaining gap to a human. Speaker info: - https://x.com/bclavie - https://mixedbread.com - https://ben.clavie.eu Timestamps: 0:00 - Coding agents are a special case of knowledge work 3:48 - Why code is unusually easy to search 5:05 - Thirty days, and when one clue means four things 5:55 - The tool loop and the organization loop 8:06 - Tooling is never a neutral add on 9:22 - A leaderboard, and two different BM25s 11:14 - How a law firm breaks down a case 12:42 - When even a human hits the tool ceiling 13:31 - Searcher agents, memos, and the oracle gap 15:13 - Design for knowledge work, not for programming --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/O84lhGc1OOI-youtube-976782f116151e7fb982 Join our upcoming AI Engineer conferences (see https://ai.engineer landing page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS 2026 in Sydney --- canonical: >- https://liquidgold.cc/channel/aidotengineer/explore/video/if-we-want-them-to-do-knowledge-work-design-them-as-de478072d38e markdown: >- https://liquidgold.cc/channel/aidotengineer/video/if-we-want-them-to-do-knowledge-work-design-them-as-de478072d38e.md llms: null parent: 'https://liquidgold.cc/channel/aidotengineer.md' channel: aidotengineer updated: '2026-10-23T22:21:07.906Z' language: null children: [] videoId: O84lhGc1OOI durationSeconds: 1075 publishedAt: '2026-09-16T15:00:39.000Z' searchChunks: - ordinal: 0 text: null startSeconds: 24 endSeconds: 230 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=24' heldReason: rights - ordinal: 1 text: null startSeconds: 24 endSeconds: 230 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=24' heldReason: rights - ordinal: 2 text: null startSeconds: 230 endSeconds: 242 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=230' heldReason: rights - ordinal: 3 text: null startSeconds: 242 endSeconds: 306 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=242' heldReason: rights - ordinal: 4 text: null startSeconds: 306 endSeconds: 318 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=306' heldReason: rights - ordinal: 5 text: null startSeconds: 318 endSeconds: 363 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=318' heldReason: rights - ordinal: 6 text: null startSeconds: 373 endSeconds: 486 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=373' heldReason: rights - ordinal: 7 text: null startSeconds: 373 endSeconds: 486 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=373' heldReason: rights - ordinal: 8 text: null startSeconds: 373 endSeconds: 486 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=373' heldReason: rights - ordinal: 9 text: null startSeconds: 498 endSeconds: 565 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=498' heldReason: rights - ordinal: 10 text: null startSeconds: 576 endSeconds: 681 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=576' heldReason: rights - ordinal: 11 text: null startSeconds: 576 endSeconds: 681 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=576' heldReason: rights - ordinal: 12 text: null startSeconds: 692 endSeconds: 762 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=692' heldReason: rights - ordinal: 13 text: null startSeconds: 692 endSeconds: 762 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=692' heldReason: rights - ordinal: 14 text: null startSeconds: 773 endSeconds: 818 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=773' heldReason: rights - ordinal: 15 text: null startSeconds: 818 endSeconds: 829 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=818' heldReason: rights - ordinal: 16 text: null startSeconds: 829 endSeconds: 916 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=829' heldReason: rights - ordinal: 17 text: null startSeconds: 829 endSeconds: 916 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=829' heldReason: rights - ordinal: 18 text: null startSeconds: 927 endSeconds: 1075 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=927' heldReason: rights - ordinal: 19 text: null startSeconds: 927 endSeconds: 1075 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=927' heldReason: rights - ordinal: 20 text: null startSeconds: 927 endSeconds: 1075 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=927' heldReason: rights - ordinal: 21 text: null startSeconds: 927 endSeconds: 1075 watch: 'https://liquidgold.cc/watch/O84lhGc1OOI?t=927' heldReason: rights --- # If we want them to do Knowledge Work, design them as Knowledge Agents — Benjamin Clavié, Mixedbread An agent searching a contract for "30 days" has no way to know whether it has found a deadline, a grace period, or a retention rule, or wandered into a medication schedule entirely. Benjamin Clavié uses that to explain why coding agents turned out to be the easy case. Code carries durable cues: identifiers, file paths, method definitions, things you can literally grep for. It also comes with narrow instructions, because nobody asks a coding agent to invent a paradigm, they ask it to close a ticket. Knowledge work outside code is the opposite on every axis. The meaning is implicit, the same string means different things in different domains, and the search has to start from an intent the agent must reconstruct on its own. His argument is that the industry generalized from the one domain that least resembles the rest. The good news is that humans solved this already, repeatedly. Clavié traces a tool loop running from the first catalogue of the Library of Alexandria through bibliographies to search engines, and an organization loop running from the single gifted polymath through monasteries and universities to the modern specialist firm. They are really one loop: better tools create new roles, new roles generate knowledge, new knowledge demands better tools. His benchmark numbers make the practical version concrete. A badly tuned retrieval tool lands near 60 percent accuracy, which nobody would trust enough to keep using. A well tuned one reaches the same ceiling as its rivals while making 20 percent fewer tool calls, and adding a layer of searcher agents that return short memos closes roughly 40 percent of the remaining gap to a human. Speaker info: - https://x.com/bclavie - https://mixedbread.com - https://ben.clavie.eu Timestamps: 0:00 - Coding agents are a special case of knowledge work 3:48 - Why code is unusually easy to search 5:05 - Thirty days, and when one clue means four things 5:55 - The tool loop and the organization loop 8:06 - Tooling is never a neutral add on 9:22 - A leaderboard, and two different BM25s 11:14 - How a law firm breaks down a case 12:42 - When even a human hits the tool ceiling 13:31 - Searcher agents, memos, and the oracle gap 15:13 - Design for knowledge work, not for programming --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/O84lhGc1OOI-youtube-976782f116151e7fb982 Join our upcoming AI Engineer conferences (see https://ai.engineer landing page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS 2026 in Sydney --- Canonical: https://liquidgold.cc/channel/aidotengineer/explore/video/if-we-want-them-to-do-knowledge-work-design-them-as-de478072d38e · Channel: [AI Engineer](https://liquidgold.cc/channel/aidotengineer.md) > Search chunk text is held until public-source provenance is verified. ## Document chunks > Missing: no stored document chunks are available.