My junior sister's experience with Vibe coding these days
1. My junior sister is definitely still a beginner at Vibe coding, and she doesn't understand many programming concepts well.
A friend told me that scripts shouldn't be run locally. My junior sister downloaded Docker but still doesn't know how to use it, and has been running scripts locally😭 this level of beginner.
2. Vibe coding is very useful, helping me solve many tasks I couldn't complete before.
For example, scraping exchange information. My junior sister scraped the token price performance over the past year, which she definitely couldn't have done before.
My junior sister's mind always comes up with various ideas, but in the past, she was limited by technology. Now with AI, she dares to try any idea.
3. Before, she just had an idea and directly sent it to AI for creation.
But this isn't really a good approach; rough ideas only produce rough products.
Often, the first version of a product won't meet your needs, so try to think through the product carefully first.
Break down the product into different parts and complete them step by step.
For example, the content in the image below:
1) First, create a data scraping script to extract each K-line, including price changes and dates.
2) Then, develop an analysis tool to analyze the extracted data, as shown below.
3) Finally, build a monitoring tool to monitor real-time data and integrate the analysis data above.
4) Maybe, at the end, refine the trading strategy and develop an automatic trading tool🤔.
4. Before, my junior sister couldn't understand code, but after using Vibe coding for a while, she gradually learned a bit of coding.
Creating an automatic trading system is actually done in different modules.
So, if there are issues, you can let AI modify that part of the code alone, without changing everything (AI tends to forget and often mess up the code).
5. A single AI might not produce the best code, but the code can be iteratively corrected or modified by multiple AIs.
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
My junior sister's experience with Vibe coding these days
1. My junior sister is definitely still a beginner at Vibe coding, and she doesn't understand many programming concepts well.
A friend told me that scripts shouldn't be run locally. My junior sister downloaded Docker but still doesn't know how to use it, and has been running scripts locally😭 this level of beginner.
2. Vibe coding is very useful, helping me solve many tasks I couldn't complete before.
For example, scraping exchange information. My junior sister scraped the token price performance over the past year, which she definitely couldn't have done before.
My junior sister's mind always comes up with various ideas, but in the past, she was limited by technology. Now with AI, she dares to try any idea.
3. Before, she just had an idea and directly sent it to AI for creation.
But this isn't really a good approach; rough ideas only produce rough products.
Often, the first version of a product won't meet your needs, so try to think through the product carefully first.
Break down the product into different parts and complete them step by step.
For example, the content in the image below:
1) First, create a data scraping script to extract each K-line, including price changes and dates.
2) Then, develop an analysis tool to analyze the extracted data, as shown below.
3) Finally, build a monitoring tool to monitor real-time data and integrate the analysis data above.
4) Maybe, at the end, refine the trading strategy and develop an automatic trading tool🤔.
4. Before, my junior sister couldn't understand code, but after using Vibe coding for a while, she gradually learned a bit of coding.
Creating an automatic trading system is actually done in different modules.
So, if there are issues, you can let AI modify that part of the code alone, without changing everything (AI tends to forget and often mess up the code).
5. A single AI might not produce the best code, but the code can be iteratively corrected or modified by multiple AIs.