To say that the technology that is very hot in the IT field this year is AIGC. Among them, the leader in the field of text processing is OpenAI's ChatGPT.
A few months ago, when this wave of AI boom started, facing the excellent performance of ChatGPT, I had a premonition that this thing is definitely not just used as a simple chat question and answer tool, it must be able to be used in many scenarios Play a role.
At that time, I asserted in a fan group: This year there will definitely be a lot of tool software and plug-ins around the application of GPT in various subdivision scenarios, and there will inevitably be a large number of entrepreneurial teams rushing to this blue ocean market.
During this period of time, I have seen various applications such as AI for PPT, AI for resumes, AI for weekly job reports, and so on.
A few days ago, when I was browsing GitHub, I saw an open source project under Alibaba on the Trending homepage:
With the fame of more than 5,000 stars and Alibaba, I clicked in. This is the Chinese introduction of the project:
At first glance, it is very similar to Navicat and DBeaver that our programmers use daily, a database connection management tool.
Developed using Electron technology, the overall interface is simple, clean and refreshing.
But since there is a "Chat" in the name, things are definitely not simple. Let's see what AI capabilities it has.
Also mentioned in its introduction:
Compared with traditional database client software Navicat and DBeaver, Chat2DB integrates the ability of AIGC, which can convert natural language into SQL, and can also convert SQL into natural language, and can give developers SQL optimization suggestions , which greatly improves Personnel efficiency is a powerful tool for database developers in the AI era. In the future, even those who do not understand SQL can have the ability to quickly query business data and generate reports.
Simply put, it can automatically write SQL according to your needs, and it can also provide some optimization suggestions for your SQL.
Specifically, it is reflected in this line of buttons on the interface:
If you want to use the ability of AI, you must be connected to ChatGPT. In order to achieve this goal, this software provides two ways, one is that the user has the API Key of ChatGPT, just fill it in and use it, of course, this requires magic to go online, you know.
Another way is to pay attention to their official account, and then obtain a temporary API Key for free experience. I will not post what the official account is here, so as not to be suspected of advertising.
Experience session
Next, let's have a brief experience and explain that all the SQL in the figure below is generated by it for us. I didn't write a single SQL myself.
First let it help us create a database and a table for testing:
Next, I let it insert 100 pieces of random data into the table I just created:
Then I found that the lips of the donkey were not right with the mouth of the horse. Where did these fields come from? It is estimated that this software has not added the ability of context linkage, and can only work for a single instruction.
Let it write some simple query statements:
It's like that, but the field name here is different from the one used when creating the table above, which once again proves that it has no contextual correlation ability.
Let's try a slightly more complicated SQL:
This is also like that.
Let's ask how it can be optimized:
This answer can be used anywhere, and it answers loneliness.
The above are just some examples, and I have used it more in-depth myself. Generally speaking, this thing is not very helpful to programmers who are familiar with SQL. The usability of the things it writes is not high. If you have time to scratch your head and think about Prompt, SQL has already been written. up.
For friends who are not familiar with SQL, what it writes may be full of mistakes and mislead users.
However, this software has only been released for a short time, and it is not bad to be able to do what it is now. It takes time for any software to grow. I believe that with subsequent iterations, the experience will become better and better. Give domestic software a try support!