100% agreed. Focus on the harness and the heuristics for output quality, that way your ongoing investment isn’t going to waste and you can capitalize on model improvements by updating the harness for the better when it makes sense.
Replying to @mattpocockuk
Stop caring so much about model releases.
Focus on the harness, and improving the environment your agent operates in.
You'll find yourself far ahead of the curve.
OpenAI's 90th percentile researchers are spending $7,000 a day on tokens so yeah, recursive self improvement is real it's just happening to the invoice
Today we're releasing data on models accelerating research at OpenAI.
Recursive self-improvement could be the most important contributor to AI capabilities over the next few years, but by default it will only be seen inside a few frontier AI labs. Being transparent is more urgent than ever, so we can inform the public discussion on whether and how to pace model development. I ask other AI companies to do the same.
openai.com/index/research-ac…
.@paste_app 1/15:
A clipboard manager sounds like a tool for accountants but it's actually for literally everyone except folks opting into a subpar life experience.
.@paste_app 14/15: MCP
Again, Paste's native search is so good that I didn't miss this before I had it.
That said here's an example of the kind of ass-saving it does when I need it.
.@paste_app 15/15:
Paste is far from new. This banger came out in 2015, and it remains my clipboard goat.
If you use a different tool for the job and rip it, God bless you.
If you think you won't benefit from such a tool, still God bless you but c'mon now.












