7 pointsby thewhitetulip8 hours ago5 comments
  • dlcarrier7 hours ago
    That's the same timeline as commercially viable fusion power plants, or fission power plants in a shipping container, or off-the-shelf self-driving cars that work in non-ideal conditions without human intervention, or batteries that charge in minutes, or work well in cold weather, or don't weigh a literal ton to get a half-usable towing range, etc…
  • N_Lens7 hours ago
    Titlegore aside, we've been 6-12 months away from automating SWEs since GPT3.0 came out.
  • andsoitis7 hours ago
    Logically, that means that Anthropic will fire all their SWE’s one year from now.
    • thewhitetulip7 hours ago
      That's what he said. AI is building AI so that's where they are headed.

      Most of my devs don't write code, they review it.

      • andsoitis5 hours ago
        I would wager $1m that Anthropic will still have SWEs doing engineering work 3 years from now, if the company is still around.
        • thewhitetulipan hour ago
          Yeah. And afaik and hage used AI models, unless you use claude opus 4.5high you don't get good results
  • Juliate7 hours ago
    I expected Anthropic to be clever enough to position themselves differently from OpenAI. This sounds like a failing marketing message; their critical long-term customers are people, not CEOs.
  • sargstuff7 hours ago
    Ummm... no time line on when AI's excpected to outsource to less expensive SWEs.