This fundamentally breaks our security posture capability for harness configuration, ending our ability to prevent company policy breaking changes by man, machine, or third party repo.
We may be a minority in using /etc/opencode in this way, but it has shaken my trust that they understand security in the Ai era or the needs of enterprise users.
I'm currently running qwen and gemma models in my homelab and using five different Chinese model families in the per-token/sub plans
I trust the open models which can be poked and prodded over the closed models that are having inputs and outputs manipulated, or flat out refusing legitimate requests
model family diversity in your workflows is important, regardless of open or closed
I use open weights and harnesses not tied to a single vendor, I definitely do not touch anything the two US Big Ai labs puts out. I'm not interested in handing power and control of humans most significant technology to the same class of people that brought us the technologies that created our mess of a digital ecosystem.
> I'm not interested in handing power and control of humans most significant technology to the same class of people that brought us the technologies that created our mess of a digital ecosystem.
So you use clones and copycats instead? You don't trust Sauron's ring so you make your own ring, maybe just no ring?
10 comments
[ 0.22 ms ] story [ 27.6 ms ] threadWe may be a minority in using /etc/opencode in this way, but it has shaken my trust that they understand security in the Ai era or the needs of enterprise users.
I trust the open models which can be poked and prodded over the closed models that are having inputs and outputs manipulated, or flat out refusing legitimate requests
model family diversity in your workflows is important, regardless of open or closed
Check and mate
So you use clones and copycats instead? You don't trust Sauron's ring so you make your own ring, maybe just no ring?
Tbf this is as old as free software instead