The A.I Wars Begin

Imagine a world without AI-powered chatbots like ChatGPT, where finding answers meant looking through endless forums, browsing Stack Overflow, or spending hours searching the web. With all the innovation humanity has achieved, it’s hard to imagine going back. Today, most of us turn to ChatGPT for a quick response whether it is a coding problem or a question we’ve never encountered before.

What makes ChatGPT so popular isn’t just that it was the first large language model (which set the standard for LLMs)¹ — it’s also because of easy accessibility to the public and more importantly, free. Think: Free as in free to use, not free software.

‘OpenAI’ being Closed A.I, became a victim of their own game. Deepseek AI’s R1 model, trained entirely through reinforcement learning (RL) matched — or even outperformed—OpenAI’s o1 across multiple benchmarks. This left stakeholders and investors in the U.S. increasingly anxious, which turned the US markets into a bloodbath. Notably, NVIDIA lost 600 billion dollars², their highest market cap devaluation on a single day!

Despite being a Chinese company, Deepseek is open-source, making its underlying code freely available to everyone—and even allowing users to run its LLM locally. So, dismissing it solely because of its origin is out of the question.

It’s fascinating to see nations competing in AI, and honestly, healthy competition is a good thing. You can’t always rely on one country to lead in cutting-edge technology forever. China has proven its capabilities, and there’s no reason India, or any other nation can’t do the same.

DeepSeek is now the most downloaded A.I app across 140+ markets³, with India contributing to 15% of the total app downloads. (Nearly 2X downloads in comparison to ChatGPT’s launch)

With India gearing up to launch a homegrown AI model by the end of this year, here’s what I believe should be the absolute standard:

1. Free and Open-Source – This is non-negotiable. Accessibility and transparency will be key to mass adoption.
2. Timely Data Updates – The model should continuously collect and refresh relevant data to stay up-to-date.
3. Robust Training Algorithms – Developing strong algorithms and feeding the model with high-quality data is essential.
4. Parameter Optimization & Testing – Fine-tuning parameters and rigorously testing performance will be crucial to staying competitive.
5. Accuracy & Benchmarking – Regularly fine-tuning the model and comparing its performance against global competitors.
6. Cross-Platform Accessibility – Make the model available across devices, with a focus on affordability. Most of India prefers solutions that come at zero cost—so a mobile app is a must.
7. Developer-Friendly Ecosystem – Provide extensive documentation, APIs, and how-to guides to foster a strong developer community.

If India gets these right, it could establish itself as a major player in the AI race.