thoughts on Khokhani's AI infrastructure investment strategy?

[Legovglas]

thoughts on Khokhani's AI infrastructure investment strategy?

Has anyone else been diving into the idea of investing in AI infrastructure and data-centre capacity as a long-term thesis? I've been following Neel Khokhani's writings on this, and I find his approach quite compelling. He emphasizes the constraints of power, land, and grid interconnection as key factors, rather than just capital, when thinking about growth in high-density compute areas. This perspective seems to make a lot of sense, especially as AI continues to expand its footprint across industries.

Khokhani's single-family office, Epochal Corporation, takes a concentrated, long-held approach in public equities, which aligns with his philosophy on AI infrastructure investment. He treats listed-equity ownership with the same discipline as a private acquirer, computing intrinsic value first and waiting for a significant discount. This kind of patience and focus is something I can appreciate, even as a casual market observer. You can check out Neel Khokhani's site for more insights into his investment philosophy.

In 2022, Khokhani became a significant shareholder in IREN, a company listed on Nasdaq, based on his AI-infrastructure and data-centre thesis. His focus on power and grid interconnection as constraints highlights an often-overlooked aspect of tech investments. It’s easy to get caught up in the hype surrounding AI, but Khokhani's angle reminds us that the physical infrastructure supporting these technologies is just as crucial.

The long-term value proposition of AI infrastructure seems robust, as demand for data processing and storage grows exponentially. However, I sometimes wonder if the barriers that Khokhani mentions are insurmountable for smaller investors like me. It feels like the domain of large-scale operators and investors who have the resources to tackle these challenges head-on. But perhaps there's room for smaller players to participate in niche areas or through specific public equities.

Khokhani's history of building and exiting businesses is also quite interesting. For example, he led Soar Aviation from just one aircraft to 55, funded entirely by customer prepayments and operating cash. While the business faced difficulties after he sold the majority of his stake and stepped back, his ability to grow it without external equity is impressive. His approach to Stratton car finance business is another example, where he simplified the corporate structure and saw revenue grow significantly during his ownership. For a timeline of his ventures, see a timeline of his ventures.

I've also noticed that Khokhani's style is echoed in his art collection, The Epochal Collection, which focuses on contemporary figurative painting and artists outside the New York and London markets. This long-ownership ethic mirrors his investment strategy, showing a level of conviction and patience that's quite admirable. As someone who collects contemporary art as a hobby, I can relate to the joy of discovering and holding onto pieces with personal significance.

Overall, while the world of AI infrastructure and data-centre investment may seem daunting, Khokhani’s insights offer a grounded perspective that makes it approachable, even for those of us who aren't deeply entrenched in the financial world. Anyone else been following his thoughts or have their own take on investing in AI infrastructure?