--- schema: 1 type: video title: 'Ending AI Slop — Thais Castello Branco, Taste Labs' description: >- Thais Castello Branco's starting point is that AI is still badly behind on the subjective work, the writing and design where quality is real but hard to pin down, and that ending the slop means building data and reinforcement environments for taste. She sorts domains along a spectrum: at one end things that verify and execute cleanly, at the other pure preference with no ground truth, and most valuable work sits in between. The move that makes taste tractable is decomposition, breaking something like a brand or a page into elements that can each be graded against an original rather than judged whole. The deeper problem she names is collapse to the mean. A model optimizing for the most likely output drifts toward the average and quietly kills the creativity that good design depends on, so the goal is data and rewards that reward breaking from the obvious when the situation calls for it. Taste Labs builds this by turning expert judgment into structured, high signal preference data, pairing it with human QA where reviewers tie their commentary to specific choices, and being careful that preference signal stays rich instead of noisy. Her argument is that as more subjective domains become measurable this way, taste becomes something you can actually train. Speaker info: - https://x.com/thaiscbranco_ - https://www.linkedin.com/in/thais-castello-branco/ Timestamps: 0:00 - What ending AI slop means 1:06 - Working with frontier labs on taste 2:33 - The verifiable to preference spectrum 3:38 - Decomposing design to grade it 5:30 - Verifying whether something is on brand 7:40 - Collapse to the mean 9:18 - Shifting preference toward verifiable 11:25 - Building a preference vector 13:05 - Human QA tied to specific choices 15:00 - Training taste in subjective domains --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/lCBf9slCanI-ending-ai-slop 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/ending-ai-slop-thais-castello-branco-taste-labs-62fb738fcc0b markdown: >- https://liquidgold.cc/channel/aidotengineer/video/ending-ai-slop-thais-castello-branco-taste-labs-62fb738fcc0b.md llms: null parent: 'https://liquidgold.cc/channel/aidotengineer.md' channel: aidotengineer updated: '2026-10-24T06:29:32.058Z' language: null children: [] videoId: lCBf9slCanI durationSeconds: 990 publishedAt: '2026-07-31T19:26:20.000Z' searchChunks: - ordinal: 0 text: null startSeconds: 24 endSeconds: 68 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=24' heldReason: rights - ordinal: 1 text: null startSeconds: 79 endSeconds: 156 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=79' heldReason: rights - ordinal: 2 text: null startSeconds: 156 endSeconds: 168 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=156' heldReason: rights - ordinal: 3 text: null startSeconds: 168 endSeconds: 223 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=168' heldReason: rights - ordinal: 4 text: null startSeconds: 235 endSeconds: 332 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=235' heldReason: rights - ordinal: 5 text: null startSeconds: 332 endSeconds: 343 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=332' heldReason: rights - ordinal: 6 text: null startSeconds: 343 endSeconds: 461 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=343' heldReason: rights - ordinal: 7 text: null startSeconds: 461 endSeconds: 473 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=461' heldReason: rights - ordinal: 8 text: null startSeconds: 473 endSeconds: 560 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=473' heldReason: rights - ordinal: 9 text: null startSeconds: 560 endSeconds: 570 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=560' heldReason: rights - ordinal: 10 text: null startSeconds: 570 endSeconds: 689 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=570' heldReason: rights - ordinal: 11 text: null startSeconds: 570 endSeconds: 689 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=570' heldReason: rights - ordinal: 12 text: null startSeconds: 700 endSeconds: 790 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=700' heldReason: rights - ordinal: 13 text: null startSeconds: 790 endSeconds: 801 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=790' heldReason: rights - ordinal: 14 text: null startSeconds: 801 endSeconds: 902 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=801' heldReason: rights - ordinal: 15 text: null startSeconds: 913 endSeconds: 990 watch: 'https://liquidgold.cc/watch/lCBf9slCanI?t=913' heldReason: rights --- # Ending AI Slop — Thais Castello Branco, Taste Labs Thais Castello Branco's starting point is that AI is still badly behind on the subjective work, the writing and design where quality is real but hard to pin down, and that ending the slop means building data and reinforcement environments for taste. She sorts domains along a spectrum: at one end things that verify and execute cleanly, at the other pure preference with no ground truth, and most valuable work sits in between. The move that makes taste tractable is decomposition, breaking something like a brand or a page into elements that can each be graded against an original rather than judged whole. The deeper problem she names is collapse to the mean. A model optimizing for the most likely output drifts toward the average and quietly kills the creativity that good design depends on, so the goal is data and rewards that reward breaking from the obvious when the situation calls for it. Taste Labs builds this by turning expert judgment into structured, high signal preference data, pairing it with human QA where reviewers tie their commentary to specific choices, and being careful that preference signal stays rich instead of noisy. Her argument is that as more subjective domains become measurable this way, taste becomes something you can actually train. Speaker info: - https://x.com/thaiscbranco\_ - https://www.linkedin.com/in/thais-castello-branco/ Timestamps: 0:00 - What ending AI slop means 1:06 - Working with frontier labs on taste 2:33 - The verifiable to preference spectrum 3:38 - Decomposing design to grade it 5:30 - Verifying whether something is on brand 7:40 - Collapse to the mean 9:18 - Shifting preference toward verifiable 11:25 - Building a preference vector 13:05 - Human QA tied to specific choices 15:00 - Training taste in subjective domains --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/lCBf9slCanI-ending-ai-slop 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/ending-ai-slop-thais-castello-branco-taste-labs-62fb738fcc0b · 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.