As OpenAI works to enhance ChatGPT, adding new functionalities and capabilities with each iteration, other large language models (LLMs) are springing up. Many are hot on the heels of ChatGPT, striving to revolutionise various industries. As LLMs develop, so too do the AI tools that draw from them, each promising a new way to simplify, remove or replace processes.
Software design and coding is one of the key areas where these tools excel, and developers have every right to be concerned about the future of their role in the development process.
There exists a global shortage of qualified and reliable talent. One of the big current questions is whether AI tools are aiming to satisfy the existing demand for these professionals, or replace them altogether. In this article, we’ll take a look at some of the tools which are currently being used, and explore the implications of them for developers.
At a Glance
v0
v0 is a ‘Generative UI’ tool created by Vercel. It is able to create user interfaces and even full site designs based on text prompts. A user can simply describe their ideas, and v0 will create the desired product.
In practical terms, users of v0 can describe the site they want to build using as simple or as complex language as they like. v0 then generates open-source platform code which can be edited within the program, and copied to the appropriate development app.
Other Applications
v0 is a UI-generator first, but it has many other uses. It’s especially well-versed in tools such as Node.js, React and Next.js. Here’s an overview of some of the ways in which developers can use v0:
|
Code generation |
|
UX and UI creation |
|
Debug code |
|
Review and edit code |
|
General problem solving |
|
Best practices |
|
Developer training |
Implications
Changing the way that UI is created could alter the way that end users interact with software products forever. v0 could have massive implications for developers too, particularly those that specialise in UI and UX design.
However, some of the most important qualities of good UI design are aesthetic, and, whilst AI can mimic and combine existing styles, a human developer is likely to be required if true originality or art is the goal.
Vercel stresses that the goal of v0 is “to help developers build the first iteration of their product,” which would suggest that a human developer will still be necessary to refine and polish it. Could v0 be the future of UI design? Perhaps, but not just yet.
Cursor
Cursor is a code editor which essentially aims to enhance a developer’s coding productivity through the use of AI.
Getting started with Cursor is quite simple. It’s a remarkably quick tool to learn to use. With so many AI coding tools being released, the refreshing simplicity and intuitiveness of Cursor will likely work to its advantage.
Features
Cursor is packed with usability features that make it a valuable weapon in a software developer’s arsenal. Here are some of the key features that make Cursor an effective choice of code editor:
|
Intelligent autocomplete suggestions that save time |
|
Multiple line changes, removing repetitive work |
|
Mistake correction to ensure high quality code |
|
Built-in AI coding chatbot that can provide feedback |
|
Use AI to modify existing code or create brand new code |
A Threat or Benefit to Coders?
Software engineers and developers can rest easy for the time being. Cursor is all about assisting developers and facilitating their creations, rather than replacing them. It’s likely to prove very helpful to a wide range of developers.
We can be reasonably certain that, as AI code editors like Cursor evolve and are refined, they will replace more conventional platforms and become the new norm. However, as it stands, there are a few undeniable drawbacks to using AI for code generation and editing.
Artifacts
Artifacts is a tool created by Anthropic that makes up part of the Claude family of LLMs and tools. Claude.ai offers a work-focused AI assistant with emphasis on security, safety and accuracy. The Artifacts tool enables the user to interact more creatively and collaboratively with Claude.
The Artifacts tool basically provides a separate window from the main conversation, which can display interactive outputs. Through this window, the user and Claude can share standalone content, create drafts and refine projects.
Examples of uses for Artifacts include:
|
Streamlining your workflow and facilitating multitasking |
|
Producing and testing prototypes |
|
Creating pieces of code and text that are separate from the main chat |
|
Generate HTML, Javascript and CSS web pages |
|
Providing visualisations and separate analyses of data |
How Artifacts Will Impact Developers
Artifacts is a tool to improve the user’s ability to interact with Claude.ai. It is likely to make working with AI a much smoother process by facilitating multitasking and enhancing productivity.
This tool should be thought of as a way to communicate more effectively with AI, rather than an AI system in and of itself. For this reason, any limitations of the underlying AI system will be present. In the case of Claude.ai, this is that it tends to ‘hallucinate’, so all work needs to be checked manually anyway.
Developers should consider Artifacts a tool that is beneficial to their work process, and not mistake it for something that can actually replace the work they do.
Copilot
Perhaps one of the best-known AI coding tools is GitHub Copilot (not to be confused with Microsoft’s Copilot AI), which proudly declares itself “the world’s most widely adopted AI developer tool”.
It was developed by OpenAI in partnership with GitHub. Copilot provides developers with a virtual coding partner. The central tenets of GitHub Copilot are:
Intelligent code completion
Automatic program generation
Increased developer productivity
Is Copilot the Future of Coding?
Quite possibly, yes. GitHub Copilot is a highly effective tool with some impressive stats to back it up:
Copilot increases coding speed by 55%
1/3rd of Fortune 500 companies use Copilot
Most developers prefer Copilot to alternatives
More than 75,000 businesses use Copilot
Even so, as automated and impressive as Copilot is, it still requires developers to operate it. Perhaps one day this won’t be the case, but, for the time being, Copilot won’t be replacing any developers.
Conclusion
The majority of AI tools still require human developers to program them, instruct them, and to check their work. As with many professions across many industries, certain types of developers may one day be rendered obsolete, but that day is not today!
