We have been committed to exploring the practical application and value of#AIin the #Web3 investment field, and allowing more people to maximize information efficiency and value capture through AI technology in the era of on-chain data transparency. This is a partial case display of our first issue: previous articles, today we are going to talk about the #GPT4 king bomb function: code parser.
How useful is this code parser? In the past, GPT4 could only ask text questions in a single direction, one question and one answer, and could not do two-way interaction. Now the code parser has powerful functions: it can upload videos, audios, tables, code files, etc. Today, I will show you a hundred-fold coin project that our investment research staff spent 50 days to screen, and let GPT4 help me analyze it! Solved in 1 hour!
The first step is to feed the source data EXCEL table to GPT4 and let it learn first (because it involves relevant internal courseware materials, only part of the display is made, please understand).
The second step is to let GPT4 help me find the correlation analysis of the projects and make a heat map of the correlation variables. The closer the value is to 1, the stronger the correlation is, and the smaller it is, the weaker the correlation is. In fact, no strong correlation was found in this step. (It is very important to show the thinking logic)
The third step is to let #GPT4 help me make a bar chart visualization according to the corresponding track. The longest pink bar is the L1 public chain. This is true. In the last cycle, the projects with the most 100x coins were mainly distributed in the public chain. We call them Ethereum killers, such as#SOL#ONE #MATIC etc.
The fourth step is to classify the categories of coins with the highest probability of 100x in different tracks using pie charts. It is not difficult to find that the DEFI track has the largest proportion, indicating that the DEFI track has the highest probability of 100x coins. The next is NFT and L1 public chain. (Note: the most births and high hit rate are two concepts)
In the fifth step, we need to consider whether there is a strong relationship between the project's growth multiple and the project's market value ranking at the time. So we made data on the correlation between the multiple and the project's low and high market value rankings. Use a scatter plot!
Finally, we let #GPT4 learn autonomously and analyze the quantifiable correlation between the 100-fold increase of the coin and other variables. It gives us this conclusion! It includes: buying time & selling time (timing), lowest price, highest market value ranking, etc.
Summary: The above cases can be used not only in Web3
We also feed financial data of listed companies, and then GPT4 allows autonomous analysis: what problems have occurred in the financial data; what needs to be paid attention to; in order to improve the profit margin, what aspects should be optimized and improved, etc. Including feeding videos or pictures, through natural language, providing automatic photo editing and video editing. AI is the new productivity, Web3 is the new production relationship!
