A Meta engineer explained how to use AI in work without losing independent thinking.

For Meta software engineer Rohan Kulkarni, working with artificial intelligence doesn’t end when the workday is over.

At work, AI has significantly accelerated the process of generating ideas and creating prototypes. Projects that used to take about a month can now sometimes be completed in just two or three weeks. However, learning new technologies isn’t limited to work tasks. Every week, Kulkarni spends four to six hours exploring new developments in technology and artificial intelligence, testing ChatGPT, Claude, and Perplexity. He spends about $50 a month on subscriptions to these services.

Like many professionals in the tech industry, he continues to work with AI in his free time. Some use such tools to create their own projects, while others use them to acquire skills they don’t have time for during the workday. The reasons vary: some fear changes in the job market, others don’t want to fall behind their colleagues, and some are simply interested in the possibilities offered by new technologies.

Although AI helps complete tasks faster, it has also introduced a new challenge—learning how to use it without surrendering one’s own thinking.

The habit of constant learning

Kulkarni’s interest in technology began in childhood. He grew up in India, and his first encounter with a computer occurred around the age of five, when his parents bought him a computer running Windows 98.

In 2021, he moved to the United States to earn a master’s degree at Stony Brook University, and just a year later joined Meta as a software engineer. A recommendation and serious preparation for technical interviews helped him secure the job. Over several years, he advanced from a junior engineer to a Senior Software Engineer.

Even before the emergence of generative AI, Kulkarni independently studied new technologies and created his own projects. With the development of artificial intelligence, only the direction of learning changed—now the main focus is on AI tools.

He uses his free time specifically for such learning, without sacrificing family and personal life. Instead, he has reduced the time he previously spent aimlessly consuming content, primarily YouTube Shorts and movies on Netflix.

How to learn to think together with AI

One of the main advantages of artificial intelligence has been not only speeding up work but also changing the very approach to solving problems.

Before the advent of modern AI tools, many ideas had to be discussed with mentors at early stages of development to test hypotheses, identify weaknesses, and choose the optimal path. Now, much of this work can be done independently, using AI as a tool for analyzing and verifying one’s own ideas even before consulting colleagues.

Over time, the approach to using artificial intelligence itself has changed. Initially, Kulkarni completely handed over problem-solving to AI, expecting ready-made answers. Later, he realized that it is much more effective to first think through the solution independently and then use AI to verify assumptions, identify weaknesses, and consider alternative options.

This approach allows one to view artificial intelligence not as a universal solution to all problems, but as a tool that enhances human capabilities.

The changing role of developers requires a new type of thinking. If previously specialists performed almost all stages of work themselves, now they more often have to manage the process, analyze AI suggestions, and make final decisions. Essentially, developers are becoming more like solution architects rather than just executors.

Career advice in the AI era

Recent layoffs in the tech industry have shown that it is impossible to control everything happening in the market. Therefore, the main focus should be on the quality of one’s own work, appreciating existing opportunities, and supporting colleagues who have faced layoffs.

Another important principle is to use artificial intelligence to enhance existing skills, rather than expecting it to completely replace professional knowledge.

Even in the era of rapid AI development, the human element remains paramount. Regardless of whether one is creating AI products or personal projects, the ultimate goal is to solve real user problems.

This is why work on any project should begin not with technology development, but with understanding what value it will bring to people.

This same approach also helps in building professional connections. Instead of viewing networking solely as a way to get a new job, it is much more important to genuinely be interested in people and build long-term relationships. This makes the professional network much stronger and more useful in the future.