←新闻详情

Peter Yang更新于 10月21日 22:46

Micky:小模型处理琐碎任务,SOTA 模型负责编排审查

"The really big models can do really dumb things for minuscule tasks."

AI 导读

Rasmic 的 Micky 认为不必所有任务都用 SOTA 模型:他用 Luna 处理大量数字和费用的报税工作,并用 Sol 作为子智能体跟踪 Luna 的工作。他表示 SOTA 模型应专注于编排和审查,因为很多小模型正变得非常出色。

引用

Peter Yang

@petergyang

"One wrong assumption in [your software factory] just destroys everything." Here's my new episode with @Rasmic, where he walked me through his 4-step agentic engineering workflow to ship production apps: 1. Isolate: Give each agent its own worktree 2. Build: Create reusable building blocks 3. Prove: Make agents record proof it works 4. Ship: Get AI to…

展开完整原文

"One wrong assumption in [your software factory] just destroys everything." Here's my new episode with @Rasmic, where he walked me through his 4-step agentic engineering workflow to ship production apps: 1. Isolate: Give each agent its own worktree 2. Build: Create reusable building blocks 3. Prove: Make agents record proof it works 4. Ship: Get AI to review your PR on a loop Some quotes from Micky: "I'm not at a point where I trust the agent. It has to show evidence of what it did and why it's good." "Planning isn't the agent writing a spec and me reviewing it. It's me and the agent discussing until I know what I want. The conversation history is the spec." "Every time a new model comes out, I start fresh with no skills to see how it works. Then I add back my software factory." 📌 Watch now: https://t.co/ju9aMQKaCE Thanks to our sponsors: @WisprFlow: 4x faster than typing with your voice https://t.co/oqHJ8bN3ll @RiversidedotFM: All-in-one AI studio for podcasts and video https://t.co/uWnS6aiMPE

查看原帖

正文

当前显示原文,中文全文译文尚未就绪。

"The really big models can do really dumb things for minuscule tasks."

From Micky (@Rasmic):

"There's no need to use state of the art models for everything. I had some tax work to do with a lot of numbers and expenses to crunch. Luna just blew through that, and I used Sol as a subagent to track Luna's work.

State of the art should probably stick to orchestrating and reviewing, because a lot of the smaller models are getting really, really good."

📌 Watch Micky break down which models he uses here: https://t.co/ju9aMQKaCE

来源

Peter Yang