You’re a fast-moving agency with big dreams and bigger hustle. You already use AI to help research trends, ideate concepts, write copy and design assets. But do you code with it? 

“Agencies need to start with developers to drive success in the age of AI,” says Shuyin Zhao, VP of product at GitHub Copilot, an AI developer tool that helps you write code faster. It’s about empowering your developers with the right tools so they can create new opportunities that, in turn, empower the rest of the world, she says.

But like all things AI, it can be hard to know where to focus your energy, which tools to select and how to implement new processes. Ahead, Shuyin Zhao and Yoav Abrahami, Wix’s head of code, share how you and your dev team can tap into the massive opportunities AI coding assistants present.

Prime your devs to adopt AI

“The speed of innovation is astonishing, and new things are emerging every day that challenge the previous day’s way of thinking and doing, so it’s really all about the developer’s willingness to learn new skills and adapt to these new technologies,” says Shuyin. 

That means empowering your devs with new tools and training sessions. Motivate your teams to do their best work by walking the talk; make sure your leadership is actively discussing, using and learning about the latest in AI. Consider delegating an ‘AI owner’ within your agency who is responsible for researching and disseminating new AI tools, trends and practices. 

To Shuyin’s point, new developments in AI are launching at an unprecedented pace (in fact, a whopping 70% of professional developers are predicted to use Al-powered coding tools by 2027, according to Gartner). So, sharing information and learnings as a team, in real time, will be crucial to your agency’s success. 

Creative director at Ogilvy Paris, David Raichman, asserts that as AI makes headlines on a weekly basis, “you have to ask yourself, ‘am I going fast enough?’” Open your lines of communication to reach those ‘aha’ insights sooner. Read more about how the team at Ogilvy Paris adopts new technologies like AI quickly and strategically.

Help your devs choose the best AI coding tools for the job

Some things you only learn by doing, but you should still guide your developers to consider several core questions when deciding which AI coding tool to use. Specifically, ask them to consider:

• How accurate and reliable are the AI code assistant's code suggestions and corrections?

• What programming languages and technologies does the AI code assistant support?

• What level of support and documentation does the AI code assistant provider offer? 

• What is the cost structure and ROI of implementing the AI code assistant? 

• What security measures are in place to protect sensitive code and data? 

• What integration options are available with existing tools and workflows? 

• What is the scalability of the AI code assistant solution? 

From there, weigh your options. Of course, you can always default to ChatGPT to write code, but consider the alternatives. GitHub Copilot increases developer productivity and accelerates software development by speeding up the coding process up to 55%, according to the company. A large differentiator between Copilot and its competitors is the emphasis the team at GitHub places on responsible AI. “Naturally, the integration of AI into the development raises ethical concerns, such as algorithm bias and data privacy, so from day one we’ve taken measures to ensure Copilot’s outputs are fair, transparent and accountable,” says Shuyin.

Another tool created specifically for web creation: the Wix AI Code Assistant, an intelligent coding companion built into the new Wix IDE on Wix Studio. The AI Assistant generates tailored code snippets and troubleshoots bugs in real time to help you ship higher quality sites more quickly. “AI is changing the coding landscape and allowing developers to work more efficiently,” Yoav Abrahami notes. “That means freeing up time so they can focus on solving bigger problems.” (Read more about Wix Studio’s AI capabilities.)

Rework your processes with an AI-first mentality

The opportunities AI presents for developers to work efficiently at scale is “actually mind-boggling,” Shuyin explains. Not only is code a major differentiator if you’re looking to score more clients, but with AI, you can also accelerate your dev velocity. 

Of course, your devs can use the AI code assistant to generate functions, but they can also ask it to check why some code isn't working, or explain segments of code from tutorials to learn how to integrate new concepts within your work. Developers can even leverage assistants to provide feedback on another teammate’s work, in order to improve the way humans write code too.

“It’s not guaranteed that your AI assistant will always output the right code—which is why you still need your developers at the wheel to steer each project. However, you can always iterate on the outputs to get closer to your intended result,” says Yoav. 

Go from writer’s block to code blocks in seconds when your devs learn to work with their AI assistants. Get them to document their interactions as they play around with AI to share their findings with their team(s).

Delegate the tactical to AI, then hone in on the strategic

To be clear, AI can’t and won’t replace your developers. “There’s a misconception that AI is coming for everyone’s jobs, but we’ve found that it has allowed developers to delve into more creative work, which, in turn, makes them more fulfilled,” says Shuyin. “This ultimately yields a secondary effect of employee retention, as you give developers greater autonomy and responsibility.”

You’ll still need developers to round out machine errors and determine which code blocks work best, but the function of coding (and therefore the role of a developer) will change from creating to curating.

“Think of AI like a coworker at your company; you’re not competing, you’re collaborating,” says Yoav. ”You assign it a task, you trust it to do its job, and then you hand it off to someone else for editing and quality assurance.” 

The combination of AI doing the coding and humans overlooking the process allows your dev team to be more agile. AI serves to augment (not replace) strategic vision and human creativity.

“Now more than ever, developers need to work closely with strategists, engineers, UI/UX designers and other domain experts to bring AI solutions to life that address more complex problems,” Shuyin says. “There are so many ways for developers to innovate today. Creativity and collaboration are even more important than before.”