Can AI Find India's Missing Children?

Hi everyone,

Welcome back to the second edition of "Nothing Artificial About It," NDTV's weekly AI newsletter.

AI had a noisy week, but that's nothing new. Before we get to the noise of OpenAI's new launches and all the drama around regulating AI, let's spotlight the "good AI" right here in India.

And don't forget to check out this week's AI lifehack, which can breathe life into your writing with instant infographics. It's one more way AI is making our lives easier - when it's behaving, that is.

A few weeks ago, Sahid SK, an AI artist and creator based in Arambag, West Bengal, teamed up with the Rajasthan police to help find missing children. Sahid is also the co-founder of Megalodon, a creative agency.

When the cops from Rajasthan approached him about recreating the old, worn-out photos of the missing children, the opportunity to make AI work for people was evident. Using AI tools, the images of the kids could be recreated, imagining how they would look today, helping people identify them years after they went missing.

"I was given old, faded pictures where you could barely see the children's faces. I reconstructed their faces using AI tools and age-progression technology to estimate their current appearance. It was challenging, but knowing that this could help reunite families was all the motivation I needed," Sahid said in a LinkedIn post.

Earlier this year, in January, Delhi police used AI tools to not just identify a dead victim's body but also find his killer. As reported by NDTV's Mukesh Singh Sengar, on January 10th, the body of a young man was found near the Geeta Colony flyover in east Delhi. His face was beyond recognition, so the cops reconstructed his image using AI and printed over 500 posters circulated across the locality. A man came forward, recognizing Hitendra, giving clues for further investigation, and finally nabbing the culprits.

The intersection of AI and crime is mostly scary. But in a very tiny way, it also promises to solve crimes, not just enable them. As I wrote in my column, AI To Solve A Murder - Future Of Crime Fighting, earlier this year, sometimes AI surprises us in good ways.

Indians in AI: Ashish Vaswani and Niki Parmar, the architects of modern AI

When I first read the title of the paper, "Attention Is All You Need," I thought it was snappy. Usually, the scientific papers have really boring titles, almost pushing layman readers like me away. The paper has provided building blocks for some of the most popular AI applications today, from ChatGPT and Bard to even Google Search and Translate.

Niki Parmar completed her Bachelor's in Engineering at the Pune Institute of Technology from 2008 to 2012 and then spent years working at Google Brain.

Ashish Vaswani worked as a scientist at the Google Brain.

Niki and Ashish are now the co-founders of an AI startup, Essential AI.

If you want to learn more about how the top AI scientists at Google shaped the building blocks of modern AI, please read the following stories. I loved reading them.

Transformers: The Google scientists who pioneered an AI revolution

8 Google Employees Invented Modern AI

Digital Afterlife

Now, let's shift to a story that left me with mixed feelings - promising and slightly unsettling. I recently read an article by BenjEdwards on Ars Technica titled "My dead father is 'writing' me notes again." Benj used a Flux AI model to recreate his late father's handwriting. He generated new notes in his father's unique script by training the AI on his dad's old journals.

While this use of AI fulfills a deep emotional need, it also opens the door to potential misuse. If we can replicate someone's handwriting so easily, what's stopping bad actors from forging documents or perpetrating scams?

The same technology that brings comfort could be weaponized for deception - even used in malicious activities like ransomware attacks.

Inspired by Benj's experience, I'm trying out Flux from Black Forest Labs myself. I've started uploading some of my late father's handwritten notes to see how it goes. I'll share more about this personal journey in the upcoming editions of the newsletter.

If you're anxious about managing your digital afterlife, please read Shadma Shaikh's NDTV column, Tech's Final Frontier: Death-Tech In India. It provides an excellent checklist to follow.

Now, more good before it gets ugly.

AI Lifehack of the week

Let's dive into something fascinating and a bit chaotic - the world of AI tools we can use in our lives and work. I've always struggled to create good visual breaks for my stories. Sometimes, I get it right, but mostly, it's a hustle. So, when I discovered Napkin.ai, I was genuinely impressed.

A huge shoutout to my good friend Jayadevan PK for introducing me to the tool.

To find out how it works, I fed napkin.ai the passage below:

AI, a frenemy

While this use of AI fulfills a deep emotional need, it also opens the door to potential misuse. If we can replicate someone's handwriting so easily, what's stopping bad actors from forging documents or perpetrating scams? The same technology that brings comfort could be weaponized for deception - even used in malicious activities like ransomware attacks.

This duality highlights why AI often feels like a "frenemy." It's capable of incredible good, yet it carries risks we can't ignore. Reading a handwritten letter from someone no longer with us touches the heart, but it also raises ethical questions about consent and authenticity.

Within seconds, Napkin.ai transformed this text into a clear chart that visually captured the complex emotions and ethical considerations surrounding AI. It was like working with a professional designer at my fingertips. The tool generated multiple infographic options, and I could pick the one that best conveyed the message I wanted to share.

News and Trends

1. OpenAI Releases New Model With Reasoning Capabilities

Expect ChatGPT to be reasonable now, but the early signs show a long way to go. Reasoning with GPT is still a paradox while using the new model o1, especially if we expect it to perform beyond the "demos."

2. Lenovo To Make AI Servers In India

The company said Tuesday it has started making large, powerful computers in Pondicherry, southeastern India, moving beyond products such as laptops and smartphones. The Chinese company will also build out its facilities in the Bangalore region, including a research lab focused on artificial intelligence.

3. Axios Says AI Will Boost the Global Economy By A Cumulative $19.9 trillion by 2030

Too many voices and numbers are focused on AI's economics and how it will create an obscenely prosperous world.

4. Sergey Brin On Google's New Bets and Generative AI

This YouTube interview with Sergey covers several interesting questions, including his return to Google, the future of the human-AI intersection, Google's product culture, and so on.

5. Italy Tests AI-Assisted Teaching In Schools To Boost IT Skills

AI tools on classroom tablets and computers will act as virtual assistants that can make learning easier for students.

Thanks again for reading and subscribing to this newsletter. Please write to me if you want us to spotlight an important AI story, profile an Indian doing significant research, or use AI to change the way we live and work.

(Pankaj Mishra is the AI Editor at NDTV. He shapes stories and conversations that help explain AI in India and how it influences people's lives and work)