Recently, the AI field, led by ChatGPT, has been stealing the spotlight. The daily emergence of new information reminds us of the early days of smartphones. In this article, we will consider how front-end development will change in the future, based on the current state of the AI field. This includes some predictions, so there’s a chance we could be wrong.
GitHub’s CEO predicts that 80% of coding will be AI-generated in the future. Indeed, the power of GitHub Copilot is evident when you experience it. It suggests code on a level that’s different from traditional static analysis-based input completion.
It’s well known that it can generate code based on comments. However, it can also suggest code from function names, generate items for switch statements, and much more. Amazon has also released Amazon CodeWhisperer, a similar service, and it’s expected that the number of players in this AI-based coding support will continue to increase.
GitHub is also developing GitHub Copilot X, an evolved version of Copilot. This is based on GPT-4 and has added dialogue functionality. It can find and suggest fixes for bugs in the code, and also generate test code.
Using ChatGPT, you can generate code for Vue or React, and GPTApp has made this into a service. If you input requirements, it generates a web application to match them.
While currently, it can only create simple applications, it’s expected that in the future, the generation of these base applications will also be AI-based. If we could also define the DB schema at the same time, it seems like we could generate code including an admin screen and necessary validations.
There are often times when you hit a question while coding, or encounter an error and have to search the web for a solution.
ChatGPT is also used in these situations. By asking the right question, it often generates code that fits your purpose.
You can also write the error message as is and ask for a solution. With web search, you have to find the answer yourself from many results, but the advantage of AI is that it gives a personalized answer.
The testing process tends to be left to the later stages of a project, and the time allocated for it gets cut as the project gets into full swing. That’s where AI comes in.
The aforementioned GitHub Copilot X can generate unit tests and such based on existing code. Other tools like mabl and MagicPod use AI during testing to realize systems that are resilient to design changes and can extract screen items.
When you develop a web application, deciding where to deploy it is a problem. Moreover, if you have to build the infrastructure yourself, it can take a considerable amount of time if you’re not used to it.
Pulumi AI generates infrastructure for various public clouds using natural language. You can freely specify for AWS, GCP, etc. This should greatly reduce the work involved in setting up an infrastructure environment.
The development environment has drastically changed in recent years due to the advancement of AI. While we’re moving towards a world where we don’t have to code, programmers have to guarantee the operation of code generated by AI. Just because it’s AI-generated doesn’t mean you can shirk responsibility.
While the amount of coding is decreasing, programmers are expected to maintain a high level of quality. It’s a very difficult challenge, but it’s already being put to practical use, and it’s expected that the evolution won’t stop. Let’s explore ways to utilize it and keep trying.