{
  "blyg": "0.3",
  "id": "62qkaj2xznf78xq8fba3q7fa6s",
  "kind": "fragment",
  "origin": "https://www.msweet.net/notes/",
  "page": "174-bronze-seals",
  "author": {
    "name": "Matthew McDowell-Sweet",
    "url": "https://www.msweet.net/"
  },
  "created": "2026-11-15T14:21:39Z",
  "updated": "2026-11-15T14:31:50Z",
  "version": 2,
  "content_md": "Agents *may* be [the platform era](https://en.wikipedia.org/wiki/Platform_economy) on steroids. Think about what that means for corporations. One logical endpoint is the last company: a singular firm whose agents do the work that used to need hundreds of thousands of businesses.\n\nEven if Anthropic doesn’t get us to the end of corporations, [Fukuyama](https://en.wikipedia.org/wiki/The_End_of_History_and_the_Last_Man) style, the direction is clear. We won’t have big tech anymore. We’ll have *mega tech*.\n\nThe platform era worked because it was ruthlessly effective: aggregate demand, drive marginal costs towards zero, and let [network effects](https://en.wikipedia.org/wiki/Network_effect) do the rest. The dangers came bundled with that success: [lock-in](https://en.wikipedia.org/wiki/Vendor_lock-in), [rent extraction](https://en.wikipedia.org/wiki/Rent-seeking) once competitors were gone, single points of failure, and power concentrated in a few boardrooms.\n\nAgents amplify all of it. Hardening against that looks like open protocols over closed APIs, [data portability](https://en.wikipedia.org/wiki/Data_portability), interoperable agents and [antitrust](https://en.wikipedia.org/wiki/Competition_law) that moves at the speed of software. For corporations, the counter-move is just as clear: own the orchestration layer, the trust and the distribution, and they’ll capture even more value than the last era did.\n\nSo the question is: what are the alternatives? The same agents that make mega tech possible also put enormous leverage in the hands of tiny teams, co-ops and open communities. A few counterweights they can reach for:\n\n- Build on open protocols and open-weight models, so no single provider can switch one off or raise the rent.\n- Pool agents and data in co-ops, so members share the scale advantages instead of renting them.\n- Make portability a feature: let users leave with their data and their agents, and win them on trust.\n- Go deep on niches the mega firms find too small or too local to bother with.\n- Federate rather than aggregate, so many small services can interoperate like one big one.",
  "content_html": "<p>Agents <em>may</em> be <a href=\"https://en.wikipedia.org/wiki/Platform_economy\">the platform era</a> on steroids. Think about what that means for corporations. One logical endpoint is the last company: a singular firm whose agents do the work that used to need hundreds of thousands of businesses.</p>\n<p>Even if Anthropic doesn’t get us to the end of corporations, <a href=\"https://en.wikipedia.org/wiki/The_End_of_History_and_the_Last_Man\">Fukuyama</a> style, the direction is clear. We won’t have big tech anymore. We’ll have <em>mega tech</em>.</p>\n<p>The platform era worked because it was ruthlessly effective: aggregate demand, drive marginal costs towards zero, and let <a href=\"https://en.wikipedia.org/wiki/Network_effect\">network effects</a> do the rest. The dangers came bundled with that success: <a href=\"https://en.wikipedia.org/wiki/Vendor_lock-in\">lock-in</a>, <a href=\"https://en.wikipedia.org/wiki/Rent-seeking\">rent extraction</a> once competitors were gone, single points of failure, and power concentrated in a few boardrooms.</p>\n<p>Agents amplify all of it. Hardening against that looks like open protocols over closed APIs, <a href=\"https://en.wikipedia.org/wiki/Data_portability\">data portability</a>, interoperable agents and <a href=\"https://en.wikipedia.org/wiki/Competition_law\">antitrust</a> that moves at the speed of software. For corporations, the counter-move is just as clear: own the orchestration layer, the trust and the distribution, and they’ll capture even more value than the last era did.</p>\n<p>So the question is: what are the alternatives? The same agents that make mega tech possible also put enormous leverage in the hands of tiny teams, co-ops and open communities. A few counterweights they can reach for:</p>\n<ul>\n<li>Build on open protocols and open-weight models, so no single provider can switch one off or raise the rent.</li>\n<li>Pool agents and data in co-ops, so members share the scale advantages instead of renting them.</li>\n<li>Make portability a feature: let users leave with their data and their agents, and win them on trust.</li>\n<li>Go deep on niches the mega firms find too small or too local to bother with.</li>\n<li>Federate rather than aggregate, so many small services can interoperate like one big one.</li>\n</ul>",
  "content_hash": "sha256:973343decc67d15cd5968d5571b368bab0d17853236f0974c1b605d8afa6610b",
  "media": [],
  "changelog": [
    {
      "version": 1,
      "at": "2026-11-15T14:21:39Z",
      "note": null
    },
    {
      "version": 2,
      "at": "2026-11-15T14:31:50Z",
      "note": "added platform-era dangers, counter-moves, and a bulleted list of alternatives"
    }
  ]
}
