Breaking: India's Union Cabinet has approved a Rs 10,372 crore AI mission, aimed at creating computing infrastructure in public-private partnership mode.

• Here are 8 things you should know about the ambitious mission.

One quick thing: Sarvam AI, the startup building AI models for Indian languages, will launch its first commercial product in the next 6-12 months

• It is building a tool that can help startups and businesses with voice-related experiences like customer support.

In today’s newsletter:

• India's fund of funds scheme has minted 18 unicorns
• Can Flipkart be third time lucky
• LLMs or ESG: Tech firms strive for balance

P.S. Making bed as a competitive sport? Scroll below for more deets!

Was this newsletter forwarded to you? You can sign up for Tech3 here

A government scheme, unveiled in 2016, has indirectly given rise to 18 unicorns, and resulted in investments in 969 unique startups in eight years.

The Fund of Funds for Start-ups (FFS), operated by Small Industries Development Bank of India (SIDBI), has become a popular backer of the Indian startup ecosystem.

• The scheme has committed Rs 10,229 crore investment in 129 alternative investment funds (AIFs), as of December 2023.

• In turn, AIFs have mobilised Rs 17,534 crore investments in 969 unique startups

18 unicorns, including ShareChat, Lenskart, Digit, Acko, Unacademy, Slice, DealShare, CureFit, Spinny, and Mamaearth, have been indirectly funded via this scheme.

Of the total startup investment via FFS, 11 percent went into the healthtech segment and 6 percent in the agritech space.

• Going beyond the top startup hubs, about 129 startups beyond Tier 1 cities received Rs 1,590 crore investment, to date.

The interim Budget of 2024-25, however, had brought down the yearly allocation for the scheme by 18 percent to Rs 1,200 crore. 

Read full story

One of the biggest surprises in India’s tech ecosystem over the last two years has been the resilience of quick commerce companies. 

The general sentiment was that the segment would crumble under the pressures of profitability. But it has held up quite well so far.

No wonder e-commerce major Flipkart wants to have a go at fast deliveries. For the third time!

• The Walmart-owned firm had shuttered a service called Nearby in 2016 and Flipkart Quick in 2022 — both of which wanted to deliver in 60-90 minutes.

• It is reportedly creating a dark store network and building up an assortment that is wider than incumbents like Zomato's Blinkit, Swiggy's Instamart and Zepto.

There seem to be three major factors playing behind Flipkart’s decision to enter 15-20 minute deliveries.

• The incumbent quick commerce players are growing fast and eating up significant chunks of the e-commerce pie, according to FMCG industry executives

• The segment has demonstrated that unit economics are improving and there's a visible path to profitability

• Flipkart's management wants to double down on making its existing customers order more frequently rather than attracting new users.

Flipkart CEO Kalyan Krisnamurthy recently gave us an indication of the company's strategy going forward.

"If you just look at the gap between the people who bought once, which is let's say 500 million versus people who buy monthly, there is a massive opportunity there. A lot of our focus goes towards bridging that gap," he said.

"It’s a velocity problem. The question is, earlier you were buying two times monthly, now are you buying four times? Next year are you buying six times?," he added.

Read the interview

Generative AI, which is currently taking the world by storm, is known to be an energy guzzler. This has put tech firms in a dilemma as they look to cash in on the nascent technology while also balancing their environmental goals.

Tech giants such as Microsoft and Google aggressively building large language models (LLMs), coupled with the rapid demand for generative AI solutions from large IT firms such as TCS and Infosys, is straining their sustainability efforts.

• Any AI image generated through an 'efficient' LLM can consume energy equivalent to a fully-charged smartphone.

• The least efficient LLM can consume power equivalent to over 950 fully-charged smartphones for the same task.

Tech companies are now working to find solutions to reduce the carbon footprint.

How? By deploying energy-efficient chips, developing improved architectures, and promoting green-coding practices.

Experts say that tech companies, especially the large ones, are going through increased scrutiny in this age and era of the internet. Hence, they are struggling to balance between profits and sustainability.

• Tech Mahindra's Jagdish Mitra told us that even though every business wants to monetise Gen AI, it will not come at the cost of the environment.

Dig deeper

• More than 100 top AI researchers have signed an open letter calling on AI firms including OpenAI and Meta to allow them to study their systems. They allege that opaque company rules are preventing them from safety-testing tools that are being used by millions of consumers.

• AI Image creation tools from companies including OpenAI and Microsoft can be used to produce photos that could promote election or voting-related disinformation, despite each having policies against creating misleading content, researchers say.

• A former Google engineer has been indicted in California for stealing AI secrets to benefit two Chinese companies.

Making your bed: Either you enjoy it or find it a tiresome chore. But what if it is made into a competition?

Yup, you heard it right! There are bed-making competitions held in countries such as Japan, China, and the United States, where participants compete with one another in front of a judge to see who can make the neatest bed in a quickest manner.

Here is a video of one such event