{
  "schema_version": "1.1",
  "id": "s8:https://www.youtube.com/watch?v=k35LeKZEhiE",
  "slug": "learning-on-the-job-the-future-of-post-training-raymond-feng-applied-com-17u0m7t",
  "url": "https://feed7.dev/p/learning-on-the-job-the-future-of-post-training-raymond-feng-applied-com-17u0m7t",
  "title": "Learning on the Job: The Future of Post-Training — Raymond Feng, Applied Compute",
  "why_included": "Training on a production harness could adapt an agent to real workflows, but non-replayable interactions break common RL assumptions and make feedback harder to turn into updates.",
  "summary": "Replayable synthetic environments support **GRPO** by running several trajectories from the same initial state, grading them, and adjusting model weights toward the better paths.",
  "practical_implication": "For a custom agent, preserve traces and feedback at the model boundary. A **bring-your-own-harness** setup can train on the actual workflow instead of a brittle simulation, even when the training team cannot inspect its source.",
  "agent_context": "Replayable synthetic environments support **GRPO** by running several trajectories from the same initial state, grading them, and adjusting model weights toward the better paths.\n\nFor a custom agent, preserve traces and feedback at the model boundary. A **bring-your-own-harness** setup can train on the actual workflow instead of a brittle simulation, even when the training team cannot inspect its source.\n\nProduction interactions may be non-replayable and off-policy, so conventional comparative rollouts no longer fit. **Self-distillation**, automated trace curation, and qualitative-feedback ingestion remain open research directions rather than general solutions.",
  "source": {
    "name": "AI Engineer",
    "url": "https://www.youtube.com/watch?v=k35LeKZEhiE",
    "published_at": "2026-07-31T22:30:06.000Z"
  },
  "source_class": "video",
  "content_type": "Video",
  "layer": "agent",
  "domains": [],
  "topics": [
    "harness-engineering",
    "agent-reliability",
    "tool-use"
  ],
  "verification": {
    "status": "source_linked",
    "label": "Source Linked",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [
    "Production interactions may be non-replayable and off-policy, so conventional comparative rollouts no longer fit. **Self-distillation**, automated trace curation, and qualitative-feedback ingestion remain open research directions rather than general solutions."
  ],
  "connected_context": {
    "meaning": "This makes production traces and harness feedback potential post-training data, extending self-improvement from observability into weight updates on the real workflow. Replayable environments fit comparative RL, but live interactions often do not, so bring-your-own-harness training depends on preserving model-boundary evidence and careful curation. Self-distillation and qualitative feedback remain unresolved paths, not established substitutes.",
    "corpus_size": 318,
    "generated_at": "2026-08-01T10:08:32.212Z",
    "connections": [
      {
        "title": "Why Off-the-Shelf AI Doesn't Understand Money — Udi Menkes, Intuit",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=Owb8g3yDyzo",
        "feed7_url": "https://feed7.dev/p/why-off-the-shelf-ai-doesn-t-understand-money-udi-menkes-intuit-0y6w9rk",
        "reason": "Intuit’s verified state-action-outcome histories specify the kind of grounded production record that harness-based learning would need, while adding causal and data-quality constraints."
      },
      {
        "title": "From Signal to PR: Anatomy of a Self-Improving Agent — Jason Lopatecki, Arize",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=9HbzAWnKbo4",
        "feed7_url": "https://feed7.dev/p/from-signal-to-pr-anatomy-of-a-self-improving-agent-jason-lopatecki-ariz-1i6f0ba",
        "reason": "Arize already uses production traces to generate investigated issues and draft fixes; this signal extends the same evidence stream toward post-training model weights."
      },
      {
        "title": "Data and Environment Curation for Post-Training LLMs — Mahesh Sathiamoorthy, Bespoke Labs",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=ewtOo0scUh0",
        "feed7_url": "https://feed7.dev/p/data-and-environment-curation-for-post-training-llms-mahesh-sathiamoorth-19ab77a",
        "reason": "Bespoke Labs narrows the implied path to improvement by arguing that curated SFT may capture much of the value before teams incur the complexity of RL."
      },
      {
        "title": "Reinforcement Learning without Verifiable Rewards — Will Brown, Prime Intellect",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=AQv3qRCG6Gw",
        "feed7_url": "https://feed7.dev/p/reinforcement-learning-without-verifiable-rewards-will-brown-prime-intel-0mauwh9",
        "reason": "Prime Intellect addresses the same non-verifiable-work gap with mined traces, controllable simulators, and judge-derived rubrics, while highlighting reward hacking as a risk for the proposed feedback loop."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-07-31T22:30:06.000Z",
  "modified_at": "2026-07-31T22:30:06.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/learning-on-the-job-the-future-of-post-training-raymond-feng-applied-com-17u0m7t",
    "json": "https://feed7.dev/p/learning-on-the-job-the-future-of-post-training-raymond-feng-applied-com-17u0m7t.json",
    "markdown": "https://feed7.dev/p/learning-on-the-job-the-future-of-post-training-raymond-feng-applied-com-17u0m7t.md"
  }
}