Hindsight is brutal.
We chased smarter routing, planners, graphs for multi-model orchestration.
But the real monster was never reasoning, it was always execution.
https://t.co/dzeVQF24zp
#AgenticAI#AISafety#ExecutionControl
@_avichawla This is very cool!
I've been playing in the same space but I approached it from a very different angle. Super excited to get to work on it!
Gonna go hand in hand with this ๐
Open source primitives are not raw material for stealth rebrands.
Donโt rename them and try to sell them back as โinnovation.โ
This is happening a lot right now in agent safety, execution control, and runtime governance.
Context below ๐
#AIGovernance#AgenticAI
@WJonno77008 Exactly, control and proof shouldnโt be conflated.
Prove legitimacy once at the boundary, then get out of the execution path.
Like peas and carrots, as Forest Gump would say.
LinkedIn pitchforks are out over 'Remove AI governance from your stack' ๐ Debate's raging - mostly from people that clearly didn't read it ๐
Who's actually read the full take? Drop your side below! ๐ฅ
#AIGovernance#AgenticAI