26-05-2025
Culture, compassion, compute: Satya Nadella on what makes a generational company in the AI age
From April 20 to May 2, the world's most seasoned executives, founders, venture capitalists (VCs), innovators, and technologists converged at the Santa Clara Convention Centre in California for TiEcon 2025. The theme for this year's edition, AiVerse, explored the vast potential of applied artificial intelligence (AI) across
11 tracks
ranging from AI in mobility and entertainment to AI in manufacturing, retail and supply chains.
The highlight of day two of the annual extravaganza was the grand keynote by Satya Nadella, who expounded on the importance of cultural capital and empathy in times of revolutionary change. The Microsoft honcho was also presented with the Lifetime Achievement Award by
TiE Silicon Valley
(TiE SV) chairperson Anita Manwani and deemed 'CEO of the decade' by Naveen Chaddha, Managing Partner, Mayfield Fund.
Here are some takeaways from Mr Nadella's keynote:
On what accounts for institutional strength
'If you're an established company or founding a company, you have to come up with an idea whose time has come. In order to build that new concept, you also need to have a new capability. And in order to build
that
new capability before it's conventional wisdom, you need culture.
Live Events
Whether you're a VC, entrepreneur, executive, or an engineer, it's all the same thing. You have a cultural posture that allows you to either skill yourself up or get associated with other people with complementary skills that are needed to build something new that is needed in the world. That framework is how I evaluate where we are as a company, and where we need to go. And it's a harsh thing, because you kind of have to be in this continuous journey of renewal.'
Discover the stories of your interest
Blockchain
5 Stories
Cyber-safety
7 Stories
Fintech
9 Stories
E-comm
9 Stories
ML
8 Stories
Edtech
6 Stories
On empathy as a critical workplace skill
'If you want to learn it all and not [just] know it all, that is the foundation of a growth mindset. This means you have to have empathy more than anything else, to be able to see through other people's eyes. We think of it as a soft skill for life, but it's also a critical soft skill for innovation. Because, after all, what is innovation if not meeting the unmet, unarticulated needs of customers?
You can call this a design mindset or design thinking; but really, it's empathy. It does ultimately come down to us having the ability at scale [after] seeing the world through other people's eyes.'
On AI computing being distinct from other technological breakthroughs
'When I came to the Valley in early 1990, it felt like a golden age of systems. We're back at that again. You're adding system software, whether it is what's happening at the ASIC [Application-Specific Integrated Circuit] or chip level, system level, or with model architectures. Who would've thought two years ago that soon enough, it'd be about test times and compute?
OpenAI set the pace by their innovation, and we were obviously thrilled to partner with them. But it's also great to see what's happening in open source. You're seeing reasoning models that weren't there even just a year ago. Now they're in all models, weight, and sizes. Just yesterday, we launched five new models with reasoning.
I look at my capex budget and say, 'Whoa, what happened there?' Nevertheless, the cooling in data centres is one of the biggest challenges we have. And the point is, it's differential cooling. There's an AI accelerator rack, then there's the rest.'
On AI applications in the physical sciences
'There are three broad domains for AI application. One is cognitive and knowledge work. That's our bread and butter.
The other is physical: robotics, the autonomous domain, drones, and what have you.
The third is science. The way we have been able to deal with this domain is by doing simulations. High performance computing (HPC) was fundamentally born underneath scientific simulation. But it turns out that to simulate nature right on a Von Neumann architecture, you just have approximations. There's no such thing as a perfect simulation. That's why we are very excited about even quantum because ultimately, the quantum computer is the real breakthrough at utility scale when you really think about simulating nature.
If you want to discover a new material, AI can help speed up the HPC by coming up with candidates. So we essentially have these models that are derivatives of transformers. You don't have to start without any knowledge. There's a variety of models you can get to on foundry for different domains, biology being the hardest of them all. The rapid progress right now is in computational chemistry and material sciences.'
