I think I made one mistake with Opinion AI.
We have now published more than 150 articles, and a ridiculous number of them are about Claude. Claude Code, Cowork, Skills, MCP, agents, Projects, models, workflows. If Anthropic released something new, there is a good chance I wrote about it.
That sounds useful until I try to send those articles to someone who downloaded Claude this morning.
I did that recently and realized the problem almost immediately.
We have become very good at explaining the second floor of the building. Some people still have no idea where the front door is.
So I wanted to write one Claude guide that starts embarrassingly simple.
If you already use Claude Code, know what MCP means, or have Cowork doing real work for you, you can probably skip this one. I already wrote Full Claude Course for people fimilar with Claude at some level, here it is
This new guide is for the person who has heard the name Claude, opened the chat box a few times, maybe asked it to fix an email, and still has absolutely no idea what people mean when they talk about Projects, Skills, Cowork, Artifacts, Connectors or agents.
And there are a lot more people in that group than AI Twitter makes it look.
So if you have a friend, sibling or colleague who keeps saying, “I know Claude, I just don’t really know what to do with it,” send them this.
I think by the end, they are going to realize they were using about 10% of the thing sitting in front of them.
The first surprise is how little of Claude you are probably seeing
Open Claude and you see a box.
You type:
Write me an email.
Claude writes an email.
Read more
