{
  "id": "fd363642e7c1",
  "url": "https://www.tobyord.com/writing/half-life",
  "title": "Is there a Half-Life for the Success Rates of AI Agents? — Toby Ord",
  "note": "Frankly I’m amazed at the accuracy for these tasks at these durations. I figured it would be a 1 or 5 minute task, not 59 minutes. \n\n> For the best model (Claude 3.7 Sonnet) it could achieve a 50% success rate on tasks up to 59 minutes _vs_ only 15 minutes if an 80% success rate was required. If those results generalise to the other models, then we could also see it like this: the task length for an 80% success rate is 1/4 the task length for a 50% success rate. Or in terms of improvement: what is doable with a 50% success rate now is doable with an 80% success rate in 14 months' time (= 2 doubling times).\n\nAlso the rate of improvement is fast. ",
  "voice": "mixed",
  "saved": "2025-06-19T17:11:51Z",
  "host": "tobyord.com",
  "kind": "page",
  "tags": [],
  "record": "https://links.thingelstad.xyz/links/fd363642e7c1/",
  "markdown": "https://links.thingelstad.xyz/links/fd363642e7c1.md",
  "issues": [
    {
      "number": 324,
      "slug": "324",
      "title": "Weekly Thing 324 / Agents, Shortcuts, Joy",
      "date": "2025-06-21T11:00:00Z",
      "section": "Notable",
      "position": 9,
      "url": "https://weekly.thingelstad.com/archive/324/",
      "page": "https://links.thingelstad.xyz/issues/324/"
    }
  ],
  "connections": {
    "same_host": {
      "count": 1,
      "page": "https://links.thingelstad.xyz/hosts/tobyord.com/",
      "nearest": [
        {
          "id": "7fb796d71d71",
          "title": "Earth Restored — Toby Ord",
          "url": "http://www.tobyord.com/earth",
          "record": "https://links.thingelstad.xyz/links/7fb796d71d71/"
        }
      ]
    },
    "same_week": {
      "count": 30,
      "nearest": [
        {
          "id": "560cdb7361bd",
          "title": "Building Effective AI Agents Anthropic",
          "url": "https://www.anthropic.com/engineering/building-effective-agents",
          "record": "https://links.thingelstad.xyz/links/560cdb7361bd/"
        },
        {
          "id": "b777e4d3226b",
          "title": "Websites Are Tracking You Via Browser Fingerprinting | Texas A&M University Engineering",
          "url": "https://engineering.tamu.edu/news/2025/06/websites-are-tracking-you-via-browser-fingerprinting.html",
          "record": "https://links.thingelstad.xyz/links/b777e4d3226b/"
        },
        {
          "id": "981c54ceb3fc",
          "title": "Yancey Strickler: Forget hustle culture. Behold the Artist Corporation | TED Talk",
          "url": "https://www.ted.com/talks/yancey_strickler_forget_hustle_culture_behold_the_artist_corporation",
          "record": "https://links.thingelstad.xyz/links/981c54ceb3fc/"
        },
        {
          "id": "8b405227b581",
          "title": "Coding agents require skilled operators",
          "url": "https://simonwillison.net/2025/Jun/18/coding-agents/#atom-everything",
          "record": "https://links.thingelstad.xyz/links/8b405227b581/"
        },
        {
          "id": "68941166ba38",
          "title": "The Conversational SDLC | Joseph Bironas | Medium",
          "url": "https://josebiro.medium.com/the-conversational-sdlc-a2bd95a9ed34",
          "record": "https://links.thingelstad.xyz/links/68941166ba38/"
        }
      ]
    },
    "mentions": [],
    "mentioned_by": [],
    "saved_twice": []
  },
  "elsewhere": {
    "wayback": "https://web.archive.org/web/*/https://www.tobyord.com/writing/half-life"
  }
}