What an action-packed week in the world of AI! The gossip is going to be as interesting as SpaceX’s rocket launches.

This week would go down in history as one made for the rocket man and his rocket-AI company. Other than the IPO filing, which kept our business and markets section busy on Monday and Tuesday, Elon Musk also went shopping for a small AI startup that could theoretically help his ailing Grok get smarter with coding and steal some customers away from that smug Dario Amodei and his ‘do-it-all’ Claude.

And we see why Musk chose this week to make the Cursor-buyout announcement – it was his friendly President who held back Amodei’s most powerful model from reaching the global markets. Anthropic tried its best to remove the restrictions, but the US seems to have other intentions in mind – to keep it as an advantageous item for itself, waging discussions and debates on sovereign AI – something which India took seriously. Was it the reason behind Sarvam’s latest funding round, which saw it achieve the status of a unicorn, with HCLTech being a major stakeholder in the AI firm? 

Endless questions, but there’s more to talk about. Hence, instead of dilly-dallying here, let’s jump right to the subject matter – a roundup of all things AI this week.

And once again, this is brought to you by a gossip-loving human, not an AI running on a server from the other side of the planet. 

Anthropic’s walls, G7 anxiety, and the battle for Indian AI sovereignty

Remember when tech had borders, for those outside China? Yeah, neither do we. Following recent Anthropic restrictions on its top-tier models, the tech ecosystem in India has broken out into a massive, urgent debate over AI sovereignty. It turns out that relying entirely on the goodwill of San Francisco startups might not be the safest bet for a country of 1.4 billion people. Anthropic, however, is trying its best to unban its Fable – nobody wants to lose business from the largest economies of the world.

This isn’t just an Indian headache, though. Frontier model access and the bottlenecking of advanced AI became a headline topic at the G7 summit this week. World leaders are finally waking up to the reality that whoever controls the weights of the largest models controls the future of economic leverage. If the West chooses to gatekeep its best algorithms, everyone else is going to have to build their own fortresses.

Speaking of building fortresses…

Sam Altman’s talent heist and a peek into the ‘strategic structure’

Uncle Sam is at it again, and this time he went shopping directly inside Google’s backyard. In a stunning talent raid, OpenAI successfully hired Noam Shazeer away from Google. For those uninitiated, Shazeer is AI royalty — one of the key authors of the seminal Transformer paper that started this entire generative revolution. 

Snatching him back is a massive corporate flex, proving that OpenAI can still pull talent from Google, disrupting Sundar Pichai’s peace of mind.

But Altman isn’t just hoarding geniuses – he’s also preparing for the political and existential battles ahead. OpenAI has quietly created a brand-new ‘Strategic Futures’ policy unit and hired Dean Ball to steer the ship. As the regulatory noose tightens around AI’s neck globally, this new unit will likely dictate how OpenAI navigates the path toward AGI without getting shut down by global governments.

Now, let’s talk about Musk and an Indian-origin founder.

Elon buys Cursor, and Aman Sanger joins the ten-figure club

If you thought SpaceX only built rockets and starships, think again. In a twist that caught the developer community completely off guard, SpaceX announced it is buying Cursor – the popular AI-powered code editor that has become the absolute darling of ‘vibe coders’ everywhere. 

Why does a rocket company need a bleeding-edge coding interface? Perhaps Elon wants his Mars-bound flight code written entirely by AI-assisted hands. Also, xAI is now part of SpaceX. 

Nonetheless, the biggest winner of this cosmic buyout, however, is Cursor co-founder Aman Sanger. Overnight, Sanger found himself vaulted into the ultra-exclusive billionaire club thanks to the massive SpaceX acquisition. 

Hence, here’s a good learning point for AI engineers – if you build an AI tool that developers actually love, tech moguls will literally fly to your doorstep with a massive check.

Jeff wanted (or not?) you to pass your glass of water to AI

While Elon is busy shopping for AI geniuses, Amazon mastermind Jeff Bezos found himself in the middle of a roaring internet controversy this week over AI’s staggering water footprint. With recent UN reports warning that AI-related water consumption could equal the basic domestic needs of billions of people by 2030, the public has been hyper-focusing on the sheer volume of H2O needed to cool burning-hot GPU clusters.

While a highly controversial internet misquote had tech forums convinced that Jeff wanted humans to pass on their glass of water to sustain a future for super-intelligence, his actual stance was all about optimism. Bezos argued that the world needs to stop looking at AI’s water consumption in isolation. Compared to global agriculture, heavy manufacturing, or traditional power generation, data centers are a relatively small share of total water usage. 

Jeff essentially says that let technological innovation outrun the constraint, even if it means building massive gigawatt data centers in space over the next two decades.