IBL News | New York
Software engineers are leading the charge in adopting AI agents as coding assistants.
According to a survey by Evans Data, a research firm, nearly two-thirds of software developers already use AI coding tools. Experts say these AI agents improve developers’ daily productivity by between 10 percent and 30 percent.
These tools suggest lines of code, identify bugs, run basic tests, translate old software into modern programming language, and generate explanatory documentation. However, they still make mistakes.
The dire warnings that AI could soon automate away millions of software engineering jobs are not shared by experienced developers, industry analysts, and academics.
The New York Times summarized in an article that the outlook for software developers is more likely to be evolution than extinction.
The dominant thinking is that better tools have automated some coding tasks for decades, but the demand for software and the people who make it has only increased.
According to this view, AI will accelerate that trend by leveling up the art and craft of software design and hyper-charging productivity.
“The skills software developers need will change significantly, but AI will not eliminate the need for them,” said Arnal Dayaratna, an analyst at IDC, a technology research firm. “Not anytime soon anyway.”
The uncertainty is how fast the technology will improve and how far it can go.
Mark Zuckerberg, Meta’s CEO, has predicted that sometime this year, AI will effectively match the performance of a midlevel software engineer.
To be relevant in the future workforce, entry-level developers are taking training programs starting with AI fundamentals courses and getting hands-on experience using AI assistants to write software applications.
To be more effective, they will need to learn how to manage AI tools and cultivate creativity, critical thinking, problem-solving, communication, and empathy.
A wealth of high-quality data used to train them fuels the progress—online software portfolios, coding question-and-answer websites, and documentation and problem-solving ideas posted by developers.
Major business software firms like Microsoft, IBM, and Salesforce have jumped in to offer AI-assisted coding programs. Microsoft’s GitHub Copilot, launched in 2021, is the early commercial leader.
According to PitchBook, which tracks start-ups, investment in coding assistants reached nearly $1.6 billion in 2024, triple the previous year.
