One quick thing: Why Mamaearth is struggling
In today’s newsletter:
Event alert: CNBC TV18 & Moneycontrol will be hosting the most definitive summit on Artificial Intelligence with top global and Indian voices on November 22nd in Bengaluru. Click here to register and get access to CNBC-TV18 & Moneycontrol Global AI Conclave'.
Was this newsletter forwarded to you? You can sign up for Tech3 here
If you’re seeing more Zepto dark stores around you or are flooded with Zepto ads on YouTube, it’s because the company is now going all out.
Zepto’s cash burn has now crossed Rs 250 crore, or $30 million, a month, four people familiar with the matter told us.
Zepto is also spending heavily on poaching talent from competitors.
“It has become crazy and they are now giving hikes of up to 50-60% to poach talent”, one of the sources said, highlighting the intensive competition.
Even as Zepto is burning Rs 250 crore a month, the company has closed a Rs 2,500 crore round.
After the latest round, Zepto’s cash balance has zoomed to $1.4 billion – all of which came in the last 4-5 months.
Kunal Shah-led fintech unicorn has rolled the dice again, this time playing bulls and bears!
Cred has applied for a stock broking licence via its micro-savings and investment subsidiary, Spenny, sources told us.
Cred’s interest in broking stems from the significant revenue surge seen by majority apps in the recent years.
Cred has been strategically building its financial portfolio beyond credit cards bill payments, its foundational offering.
Not a core offering, but broking comes as a lucrative add-on and a natural extension to Cred’s aimed full stack financial play.
Stock broking could bring transaction fees, account management charges, and advisory revenues for Cred.
Will Cred be able to translate its 1.3 crore users into a solid base for the upcoming service?
For the past few years, the tech world has been gripped by the high-stakes race to develop the most advanced artificial intelligence (AI).
AI giants are facing hurdles, leading many to question whether bigger and more powerful models are the answer.
The AI industry has long followed the so-called "scaling laws," which suggest that bigger models—those with more parameters, trained on larger datasets, and supported by powerful computing—would lead to better performance. But it turns out, there are diminishing returns.
The leap from GPT-3 to GPT-4 was extraordinary, but subsequent releases, like OpenAI’s Orion and Anthropic’s Claude 3.5, have not delivered the same level of improvements.
Tarun Bhojwani, Head of People + AI argues that the real challenge is not scaling these models.
"There's no company which wins a billion dollars for having the biggest models. Companies win when their models are being used in real-life applications and they're getting paid," Bhojwani said.
Another hurdle: the data needed to train these models.
Meanwhile, developing cutting-edge models can cost hundreds of millions of dollars. As models grow larger, so do the costs.
With traditional scaling hitting a wall, companies are exploring new approaches.
First up, we have a superhero showdown with Deadpool & Wolverine!
Not enough? Then be ready to rumble with Jake Paul vs. Mike Tyson!
Streaming on Netflix at 6:30 AM IST
Note: By subscribing to Tech3 you have already made the right choice. Top it up with a premium offering, the Moneycontrol Pro Panorama, a newsletter that gives you a sharp take on macros, markets, business and finance. Sign up for Pro from this link to get this newsletter in your inbox and also a host of content enjoyed by over a million subscribers.