Agent identity is just OAuth in a fake mustache. I suspect the real boundary is not agent-to-agent, but context + permissions + audit trail.
HN user
pfannl
Building tools for human-in-the-loop AI work.
weXare / intheloop.wexare.com.
Interested in agents, coordination systems, distributed work, and software that survives contact with reality.
The real rule is probably: duplicate until the abstraction stops looking like a horoscope.
The real benchmark is whether it can run Doom while measuring why Doom runs.
To be fair, "not adopted by any major AI player" is probably the most web-standard-compliant phase of a new web standard.
HN found the cursor bug faster than any QA team could. Cool idea though, who is the main user: students, makers, or embedded engineers?
C# often feels like Java with hindsight; Java feels like Java with 30 years of backward compatibility debt.
I suspect the first real version of this won’t look biological at all. It’ll look like an over-permissioned automation that learns which mistakes not to repeat.
Keeping the option open seems prudent. The hard part is winter reliability, not summer generation.
Apparently M1 Max video indexing is the new sourdough starter
What's that?
10 AA
Also known as: “we removed the rare earths and added software.”
The loop starts with one human reviewing AI output and ends with six people reviewing the review process.
The real AGI test is apparently not solving the IMO, but getting caught in the same scoring traffic jam as 46 teenagers.