Amidst all the fear, uncertainty, and doubt caused by the impact of generative AI, GitHub has discovered something surprising. happier.
Software developers have too much work to do, says Inbal Shani, chief product officer at GitHub. Coding may be the job, but developers also write documentation, write tests, run tests, meet with internal and sometimes external stakeholders, perform code reviews, and… You need to deal with system architecture, and, oh yeah, debug existing code.
“The focus of AI from a developer’s perspective is really about productivity,” Shani says. “Developers are seeing it as a force in the industry, they understand the optionality it brings, they understand the flexibility these tools bring to their companies and the potential benefits they can derive from it. Therefore, we would like to use it.”
Industrial revolution in software development
AI is ushering in an “industrial revolution in software development,” Shani argued. That’s because AI will fundamentally change the way developers write code and the way they think about software development.
Time is just one measure of developer productivity, she said. Productivity, as measured by GitHub, is also about reducing developer load and stress.
“Once we started testing AI within GitHub, we started to see improvements in developer happiness,” she said. “As a junior developer, Copilot became a pair programmer who worked with them from the beginning while the senior developer supported them. There are more complex tests and you find yourself spending more time thinking about the documentation than just writing it.”
That’s probably why 92% of developers report already using AI, she added. In Microsoft’s January earnings report, GitHub shared that 50,000 organizations and its 1.5 million developers have adopted Copilot.
“This is probably the fastest scale adoption of developer tools and transformation in the last 70 years of code writing,” she said. “So this magical unicorn, this AI, must be doing something right. And its biggest impact is, yeah, looking at Copilot, it’s really important that developers have It’s about solving big problems, taking the burden off of them, and allowing developers to focus on what’s important, which is really the complexity of writing code and the complexity of thinking about system architecture.”
AI adaptation
Relatively speaking, AI is still in its infancy. When Shani entered her IT industry, she was an applied scientist and developed algorithms for specific challenges to solve AI. It was still a niche solution, she said. However, 2023 was a transformative year in which AI democratized.
“We’ve started bringing more AI capabilities into software development at a scale we’ve never seen before,” she said. “What we’ve seen in the last year and a half, or almost two years, is a magical black box that only certain people know how to adjust,” he said. This makes it easier to use in software development.”
AI will also change software culture, she says. In fact, that’s already the case, she added.
“If you can remove some of the stressors to significantly improve developer productivity, you create a happier, more productive environment for that organization to thrive,” she says. I did.
Just as the Industrial Revolution changed jobs, so will AI. For example, it was enough for a young developer to focus on coding and learning the organization’s coding conventions for the first few years. But AI can help bring young developers on board faster by teaching them what they need to know in terms of how the organization thinks about code, syntax, naming conventions, expectations, and guardrails. He said it could be done.
“We have already seen several customers using Copilot to help onboard new developers. [in] Here’s how you write code at your company,” Shanice said.
Changes that GitHub recognizes
Generative AI allows developers to be more creative with what they create with code. This is because developers can write programs more quickly and automate tasks that previously had to be handled manually.
“You don’t have to use Copilot to write your entire code. You can choose where you want to use Copilot, and these are usually the areas that you’re not comfortable working in,” she said. “It’s really their own choice and this is why they are pilots and co-pilots are co-pilots.”
By applying generative AI, developers can adapt it whenever and wherever they need it, or to help with tasks they don’t currently like or don’t always have time to do well. For example, she said one of the areas where GitHub is applying her Copilot is to help identify when a set of code has security issues.
“We’ll also make recommendations on how to fix vulnerabilities and which repositories to avoid because vulnerabilities have been detected. So we’re seeing a big piece of AI grow beyond the original Copilot. I understand,” she said.
Boot camps and university programs will also need to adapt, she added. For example, developers spend their first few years learning and focusing on code. Developers still need to start with the basics of coding, but are also expected to learn system architecture and design earlier. Perhaps more importantly, she added, AI tools will become a regular part of the curriculum as they become the new standard for writing code.
“We need to teach developers, front-end developers and back-end developers, how to use it,” she said. “Yes, they need to understand code. They also need to understand code…but it also moves the need for system design and system architecture much earlier in their career.”
YouTube.COM/THENEWSTACK
Technology moves fast, so don’t miss an episode. Subscribe to our YouTube channel to stream all our podcasts, interviews, demos, and more.
subscribe
