- A Meta software engineer spends four to six hours a week learning AI outside work.
- He cut back on YouTube and Netflix to make more time for experimenting with AI.
- He says the key to using AI is not letting it do the thinking for you.
For Meta software engineer Rohan Kulkarni, keeping up with AI no longer ends when the workday does.
At work, he said AI has sped up brainstorming and prototyping, with projects that once took about a month sometimes taking just two or three weeks. But keeping up with AI isn't confined to the office. He typically spends four to six hours a week learning about new developments in tech and AI, including experimenting with ChatGPT, Claude, and Perplexity. He pays about $50 a month for the three subscriptions.
Kulkarni, who's in his late 20s and lives in California, is among the many tech workers spending hours of personal time with AI long after the workday ends. Some are using it to build side projects and learn skills they don't have time to develop during the workday. Their motivations vary, ranging from fears about AI reshaping their careers and concerns about falling behind colleagues to genuine curiosity about the technology.
As AI has helped Kulkarni complete tasks faster, it's also come with a new challenge: using AI without letting it do the thinking for him.
A habit of continuous learning
Kulkarni grew up in India and said he became fascinated with technology after his parents bought him a Windows 98 computer when he was about 5 years old.
He moved to the US in 2021 to earn a master's degree at Stony Brook University, before joining Meta the following year as a software engineer. Kulkarni credited a referral and intensive interview preparation with helping him land the role. Since then, he's advanced from a new graduate software engineer to a senior software engineer.
Kulkarni said he'd been teaching himself new technologies through side projects long before generative AI became mainstream. The rise of AI has simply shifted what he was learning.
"My main motivation is that new technologies create new opportunities," he said. "I want to be well prepared to take advantage of them when they arise."
That preparation happens during his personal time, but Kulkarni said he's been able to avoid cutting back on the things that matter most to him, like time with family.
"I have reduced time spent on scrolling, consuming content without purpose," he said, adding that he's cut back the most on YouTube Shorts and Netflix movies.
Learning to think with AI
Kulkarni said one of AI's biggest benefits isn't simply helping him work faster — it's changing how he approaches problems.
Before the availability of generative AI tools, Kulkarni often turned to mentors early in the development process to bounce around ideas, identify blind spots, and think through different approaches. Today, he said AI has become a "thinking partner," allowing him to work through many of those questions on his own.
"Mentors are still important," he said, "but you can do a lot of thinking with an AI as your thinking partner before reaching out to other people."
But learning how to use AI effectively has been a process. Kulkarni said he initially "delegated thinking" to AI, giving it a problem and asking it to figure everything out for him. He's since realized a more effective approach is to develop his own ideas first and then use AI to "pressure test" them, challenging his assumptions before deciding how to move forward.
"Don't think of AI as a fix-all," he said. "It's more of a tool which is empowering you."
Kulkarni said he's grown more accustomed to this different way of thinking, but understands why the shift can contribute to "AI fatigue" among some tech workers. Before the rise of generative AI, he said, workers were often responsible for carrying out every part of a task themselves. Today, they increasingly spend time working with AI tools that can help brainstorm and evaluate ideas — requiring a different way of thinking.
"You almost think of yourself as an architect at this point," he said, "which requires a different muscle."
Career advice for the AI era
Kulkarni said layoffs across the tech industry have reminded him to focus on what he can control. Rather than worrying about business decisions, he said he tries to do his job well while supporting colleagues who have been affected.
"All I can do is do my job well, be grateful that I have a job, and then be compassionate for the people around me," he said.
Kulkarni has two other pieces of advice for tech workers. First, he said people should use AI to build on the skills they already have.
When it comes to navigating a career, Kulkarni said people should remember that, despite the AI boom, "everything is human." Whether they're building products with AI or working on side projects, he said the goal is ultimately to solve problems for other people — and keeping that perspective can help guide better decisions.
That mindset starts before work even begins. Rather than jumping straight into building something, Kulkarni said he first asks what value a project will provide to the customer.
He said the same human-focused mindset extends to networking. Rather than viewing conversations purely as opportunities to advance their careers, he believes stronger connections are formed when people take the time to get to know one another.
"Be genuinely curious to know the person and not just be there for the job," he said.
Do you have a story to share about learning AI or working in tech? Reach out to the reporter via email at jzinkula@businessinsider.com, or via Signal at jzinkula.29.
The post A Meta software engineer shares what it takes to keep up with AI — without letting it think for him appeared first on Business Insider
