Hello Betamax, For me, AI summits often feel like high-tech show-and-tell sessions with lots of flashy demos and lofty promises. So watching this year's India AI Impact Summit in New Delhi from afar was a breath of fresh air. Much of the focus was put on the infrastructure that helps companies deliver on their promises. The summit saw massive hardware commitments to back up India's software dreams. For example, OpenAI was announced as one of Tata Consultancy Services' first data center clients, coming against the backdrop of the Indian firm's 1-gigawatt AI infrastructure project powered by green energy.  Meanwhile, Anthropic CEO Dario Amodei said the Claude developer is open to working with partners in India to evaluate its models' safety and security. The company is also opening up a new office in Bengaluru. Of course, with fierce rivals like OpenAI and Anthropic sharing the same stage, there was some inevitable drama. Prime Minister Narendra Modi tied it all together with his MANAV Vision, which aims for moral, accountable, national, accessible, and valid AI. It's basically a human-centric roadmap for the AI age. As Aman Singh - co-founder of the AI-native investment bank S45 - told me, the country is moving past the pilot phase of the AI era. "India's AI opportunity will be realized not through pilots, but through disciplined deployment into the economic rails that power growth," he noted. But shifting to the execution phase is not exactly a smooth transition. While billion-dollar investments grabbed headlines during the India AI Impact Summit, one standout moment was a US$2,800 robot dog that showed some cracks in the country's AI dreams. Galgotias University presented the robot as developed in-house, but it was later discovered to be an off-the-shelf Unitree Go2, a Chinese robotic dog. Two government officials reportedly branded the episode as an "embarrassment" for India. Not a stellar moment, but there's at least a silver lining. As India pushes its AI agenda forward, the "fake it 'til you make it" culture that's been prevalent in the space is stopped dead in its tracks. As Singh puts it, "it is also important not to confuse one booth with the state of execution." He mentions that Indian teams are already shipping. Some examples he gave are vibe-coding platform Emergent, which hit the US$100 million annual run-rate mark within eight months, and diagnostic AI platform Qure.ai, which has already seen global adoption. And with these deep investments into India's AI sector, Singh says that continued "execution becomes inevitable." Miguel Cordon, journalist | 
