惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

量子位
D
Docker
月光博客
月光博客
MongoDB | Blog
MongoDB | Blog
Vercel News
Vercel News
美团技术团队
博客园 - 叶小钗
I
InfoQ
Jina AI
Jina AI
博客园 - 司徒正美
雷峰网
雷峰网
B
Blog
Y
Y Combinator Blog
A
About on SuperTechFans
WordPress大学
WordPress大学
酷 壳 – CoolShell
酷 壳 – CoolShell
大猫的无限游戏
大猫的无限游戏
Microsoft Security Blog
Microsoft Security Blog
Stack Overflow Blog
Stack Overflow Blog
腾讯CDC
H
Hackread – Cybersecurity News, Data Breaches, AI and More
Recent Announcements
Recent Announcements
V
V2EX
N
Netflix TechBlog - Medium

Sanity.io

A Board Game agent built using Sanity Context and Vercel's AI SDK | Sanity Build a prototype with Claude Code that your whole team can edit | Sanity What’s New - May 2026 | Sanity I built a London pub guide with v0 and the Sanity MCP in six hours. Here's what I learned. | Sanity Build a conference concierge with Agent Context and Anthropic | Sanity Build a content-aware Telegram agent with Vercel AI SDK and Chat SDK | Sanity How I used Agent API to generate photos for my family’s recipes | Sanity What’s New April - 2026 | Sanity Better context, better matches: An AI love story (for dogs) | Sanity How to write for an agent | Sanity Content Agent, meet Slack: AI content operations in your workflow | Sanity Structure powers intelligence | Sanity Your agent needs better content. Here's how to give it. | Sanity How to serve content to agents (a field guide) | Sanity Sanity TypeGen GA: Automatic TypeScript types for content and GROQ | Sanity Sanity is now available on the Vercel Marketplace | Sanity The logo soup problem (and how to solve it) | Sanity Content Releases: From scattered updates to coordinated publishing | Sanity What's New - February 2026 | Sanity How we solved the agent memory problem | Sanity v0 Builder Challenge: The winners | Sanity Introducing: Sanity Agent Skills | Sanity Content Agent: Days of work in one conversation | Sanity Our Sanity Values | Sanity Open Source Pledge 2025: Stepping up when it matters | Sanity v0 builder challenge: $3000 in prizes | Sanity Why AI Breaks Without Structured Content Operations | Sanity What’s New January - 2026 | Sanity BFCM 2025: What teams built when infrastructure stopped being the problem | Sanity How AI shaped holiday shopping and what it means for content in 2026 | Sanity
Learning the new skill of working with AI | Sanity
Simeon Griggs · 2025-03-08 · via Sanity.io
Yelling at software is the new way to operate things
— Even Westvang

Perhaps the most difficult thing about working with AI is filtering through the signal-to-noise ratio that currently pervades AI discourse.

Code && Content Podcast

This is a recap of one episode of the Code && Content podcast, head to the show page to find out more and subscribe on YouTube or the podcast player of your choice.

Subscribe

I'm not completely cynical about AI's relationship with my work. It is a new and amazing tool that is never going away. We're not going back. However, I have long-held reservations about its usefulness with limited human intervention.

AI is a revolutionary technology, but it's also being oversold as a one-click solution to all of life's problems. This is particularly true when it comes to conversations around replacing the jobs of creative and information workers, such as those who create content.

AI tools are legendary at creating meaningless content, popularized by the term "AI Slop."

Being too quick to create content without context or meaning is a trap I was falling into until I was corrected in a recent conversation with today's guest, Even Westvang. You can't expect great results without great inputs. So, as great as AI is at creating content, it doesn't know what you don't tell it.

AI is capable of creating unique content, but only if it is guided by someone with a unique point of view.

I've described this as "micromanaging AI," essentially having to write out many small tasks to get a good result. If, like me, you prefer to "do" rather than "plan," this can be a frustrating bit of friction when adopting AI tooling.

This conversation included many light bulb moments, but I think perhaps the best clarity I got is that working with AI tools is like learning a new skill. It's not something we should expect to innately be able to do. It actually changes our workflow. It changes how we approach creating content and performing tasks. It is a new tool and should be treated as one that requires practice to get the best results.

The pace of development in this space is so fast that there are things I would like to have included in this conversation that have changed between recording and publishing it.

For example, I think one of the best ways to work with AI is to prioritize asking it questions over strictly telling it what to do.

I've also found that using voice dictation tools like Superwhisper reduces the friction of having to type out long prompts and additional context. AI tools are extremely capable of structuring unstructured content. So, ramblings captured from your free-flowing conscious can still be processed into well authored, new pieces of content.

This blog post was dictated with Superwhisper, corrected by Grammarly, and then run through Anthropic Claude for one final cleanup.