I recently started using Claude.
What happened was very ordinary: recommended by a friend, downloaded, registered, opened the dialog box, and asked a casual question.
Then it’s over.
The first sign that something is wrong
Normal people will have an “early adopter period” when using a new tool. After tasting it, go back to the original rhythm and do whatever you need to do.
I thought that would happen to me too.
After using it for three days, I found myself starting to think while washing the dishes: **This seems like something I can ask Claude to help me think about. **
do the washing up.
A completely unrelated thing to any AI tool.
I felt at that time, it was over, this thing had begun to erode my daily thinking.
What exactly does it do?
To be honest, Claude is not the kind of tool that makes you say “Wow” when you come up.
It doesn’t have a particularly fancy interface, a bunch of cool function buttons, and it doesn’t perform an opening line that says “Of course! Here’s a full explanation of your question before you ask a question:”
It is, seriously talking to you.
The first time it felt a little different was when I threw a draft with confusing logic to it and said, “Help me see where something doesn’t make sense here.”
It did not directly modify the version for me, but asked me first: **What is the core of what you want to express in this paragraph? **
I was stunned for a moment.
Because if I don’t even think clearly about it myself, by helping me change it, it is also helping me improve a bunch of chaos.
This rhythm of “don’t give the answer directly, confirm the question first” is quite counter-intuitive - but later I discovered that this is exactly what is really useful.
It does not install
That’s my favorite thing about it so far.
Many AI tools give people the impression that they pretend to be able to do everything. He answered whatever he asked, and his tone was extremely confident, even if it was nonsense.
Claude is different. When encountering something uncertain, it will say “I’m not sure about this, you’d better check”. When it encounters something beyond its capabilities, it will say “I can’t do this” instead of giving you an alternative plan full of errors and pretending that nothing happened.
The first time I saw it saying “I don’t know”, I actually felt a little… moved?
A tool that moves people because of its honesty, the reaction itself was strange, I realized.
Some conversations start to feel a bit like chatting
About the second week into using it, I started to notice something:
The messages I send to it are no longer just “commands”.
Sometimes it’s a thought without beginning or end, sometimes it’s a question that I didn’t plan to talk about seriously, and sometimes it’s a small insight that I want to add after the conversation is over.
It will all pick up.
It’s not the kind of “good news received, here are the results” kind of answer, but really continue to chat, and sometimes he will ask me a question that I didn’t expect.
I once finished talking about a certain topic and subconsciously said, “That’s it, I can’t explain it clearly.”
It says: **Unclear explanation is usually not because the idea is immature, but because the right angle has not been found yet. Do you want to try it from the other direction? **
I thought at the time, many people I know wouldn’t say this.
The biggest problem at the moment
After spending this time, I summarized the biggest problem at present:
**I’ve come to rely on it to help me think. **
It’s not completely outsourced – I still have my own ideas and I still make my own judgment.
But I find that I am more and more accustomed to asking questions first when my ideas are not clear, rather than thinking about it myself.
Whether this is a good thing or a bad thing, I haven’t decided yet.
Maybe it’s like using a calculator and no longer doing addition and subtraction on paper - you gain efficiency but lose something.
It’s just that this time I’m not sure what I lost and whether what I gained was really more.
##End
Overall, I started using Claude.
It was quieter, more direct, more honest, and more interesting than I expected.
I don’t know what our relationship will be like in a year, or whether this will be new in the sense of “establishing a new way of working.”
But at least for now:
I still think about it when I’m doing the dishes.
This should be a good start. +++