I was so waiting for the sales pitch at the end there I was even in suspense a little
At @vercel we got to test this model before release. Best way to summarize it: astonishingly brilliant.
One interesting takeaway: it's very clear that this new generation of models can take on unprecedented bodies of parallel work.
To actually seize this intelligence, you need cloud sandboxes. Full stop. The model can delegate and orchestrate far more than a single computer can handle. Likely far more than you can even imagine it can handle.
Your bottleneck is now ① your ambition and ② your parallel computing power.
Btw, I learned from @cramforce you don't need to overcomplicate it. Install Vercel CLI, setup the Plugin and tell the agent to use /𝚘𝚏𝚏𝚕𝚘𝚊𝚍-𝚝𝚘-𝚜𝚊𝚗𝚍𝚋𝚘𝚡 skill. It'll provision and de-allocate cloud sandboxes as needed. You don't need a "multi-agent remote harness ralph orchestrator" 😆
2
15
851
My bread and butter is basically agent deployments, data platforms and automation. If I hear someone that doesn't have a decent engineering team tell me about on prem deployments in production for OS models because of "sovereignty" and "cost" I mostly know they're larping
What I still don't quite understand is a lot of these big bank roughly technical middle managers ik are enamored w/ the idea of simply inferencing stock OSS models
I don't understand why, they can't run abliterated releases or anything, these are routinely far from frontier w/ absurd upfront costs & terrible economics unless you'd like to effectively like to become an internal firmwide neocloud
theres still this false allure w/ open-source, worse than ever, but again bullish for chinese model providers again
2
51
1,704


























