Want to write a program that can run without any computer foundation? In theory, it seems to be feasible to use the power of AI.
But once you encounter an error, can AI really help you solve it?
Before writing this article, we interviewed 3 front-line programmers who rely on AI programming. We also asked them some questions, which I think it is necessary to share with everyone.
1. Can your work be completed entirely by AI?
Answer: Simple ones can, but if the work is too simple, I submit it to AI, and after AI helps me answer, the concept of error is very high. I have to tell AI my real use scenario. If I don't describe it clearly, then AI will answer the wrong question. When I finally describe it clearly, AI can indeed give a relatively correct answer.
However, I can complete such a simple job without AI.
2. What kind of program do you think AI can help you complete?
Answer: Generally speaking, AI's answer can still be used as a reference in clearer situations. But there are many situations in the actual program, and the answer of AI will depend on the way you ask the question. So far, AI is only filtering and integrating data, and the concept of error is really high.
3. Do you think that relying on AI can completely replace your job?
In response to this question, all three people said that it is not possible yet.
4. Do you think AI programming will leak company privacy?
Answer: Definitely. Unless the questions you ask avoid all the information of your company, but as long as you are not careful, AI may know your private information.
For example, when you use the advanced features of ChatGPT, you need to register an account. Once you need to use the Plus or Pro features of ChatGPT, you need to pay, and your payment information may be known to the other party. If you use the Team feature of ChatGPT directly, your organizational information may also be known to the other party.
Even if you ask very cautious questions, when asking AI questions, you may inadvertently reveal your industry.
5. What confusion do you have when using AI programming?
A: Security issues must be considered. If the code provided by AI contains the risk of SQL injection, or directly contains some implanted viruses, this is unacceptable for programmers.
In addition, if the code written by AI is too complicated and not the optimal solution, although the problem can be solved, it is not the best way to solve the problem.
At the same time, we must also consider the marginal cost of using AI. If AI is asked a question and the answer is not satisfactory, it is also a very frustrating thing.
If you happen to be a person who is curious about programming and has no basic knowledge, you might as well list some of the functions you want to do most and let AI help you do one randomly. Remember, after AI answers, you copy these codes and put them on your software compiler to see if they can run correctly?
At this time, for you who have no basic knowledge, how to put the code written by AI on your software compiler has become a new problem. For example, if I make an IOS App, which software should I install? If I want to write a website, which software should I choose to program...
Let's go back to the original question again. Is AI zero-based programming really reliable?
Originality is not easy, please contact us for reprinting. For more good articles about AI, please visit: https://iaiseek.com/news