NEWS YOU CAN’T MISS

AI Is Taking Over 1.6M Indian Voices

It’s complicated.

When I look at this woman, it’s complicated. Here’s why.

For decades, India’s call centers have powered the global customer service industry. Workers like her became the friendly, dependable voice on the other end of the line, handling tech support, billing issues, and travel bookings for companies around the world.

The formula was simple: a skilled, English-speaking workforce and low costs turned India into the world’s back office.

Now, that model is changing fast.

• AI is automating many of the tasks that once required thousands of human agents.

• In Bengaluru, startups such as LimeChat are training chatbots that can handle customer queries on their own. “Once you hire a LimeChat agent, you never have to hire again,” says co-founder Nikhil Gupta.

• Similar efforts by firms like Haptik and Reliance Jio are pushing out AI agents that can respond instantly, speak in Indian-accented English, and even calm angry customers.

For consumers, the shift is largely positive. AI delivers faster, more accurate, and personalized responses. It does not take breaks or put anyone on hold. For companies, it cuts costs and improves customer satisfaction.

But for workers like this woman, it is a turning point. India’s 1.6 million call center employees face slower hiring and growing layoffs as automation expands. The future should ... must ... still include her.

That means retraining in complex customer interactions that humans handle best, and in new skills like prompt engineering to train and work alongside AI.

AI Is Helping Build a Mini Sun on Earth ☀

What if we could trap a tiny version of the sun on Earth and use it to power everything from homes to AI supercomputers? Bear with me, there’s a bit of science in here but it’s worth it.

It’s called nuclear fusion, and Google’s DeepMind is helping make it real.

Fusion is the process that powers the sun. Instead of splitting atoms (like in nuclear plants today), it fuses them together, releasing huge amounts of energy without carbon emissions or radioactive waste.

The major challenge is keeping that reaction going. On the sun, gravity holds it all together. On Earth, scientists use giant magnets to contain the plasma: a swirling cloud of super-hot gas that can reach over 100 million degrees Celsius. 🔥 

• That’s where AI comes in: Google’s DeepMind lab has partnered with the energy startup Commonwealth Fusion Systems to build and manage a new fusion reactor. They’re using AI to control the plasma in real time.

• DeepMind’s software acts like the reactor’s brain - simulating plasma behavior and learning how to adjust thousands of settings every second to keep things stable.

When it’s switched on in late 2026, scientists hope it will be the first fusion machine in the world to give back more energy than it uses … possibly in 2027.

If fusion works, it could mean clean, near-limitless energy made from ordinary water - enough power for homes, factories, and even the computers running future AI.

Out-of-Office? Chat With Their AI Twin

Your office project is due tomorrow, but the one person who knows what’s going on is on vacation … somewhere with no Wi-Fi and a coconut in hand. 🏝️ Meanwhile, your team’s stuck waiting for an answer that isn’t coming anytime soon.

That’s the headache a new startup called Viven wants to fix. It builds a digital twin of every employee: an AI version trained on their work emails, chats, and documents, so coworkers can ask that person’s twin questions even when they’re offline.

In theory, it means no more delays, endless Slack messages, or waiting days for one approval. Just quick answers from your AI coworker’s clone. But it also raises some big questions ike how much of you your digital twin should know, and how to ensure it doesn’t divulge sensitive information.