The World Needs More AI Knowledge Diffusion

I spent the last three years in San Francisco watching the AI boom firsthand.

Everything about how work gets done changed rapidly. Companies began adopting agents that worked alongside humans in almost all critical functions like engineering, product development, operations, finance, and (ironically) HR.

At Airbnb, I myself wrote more code in a week than I had in a decade. We built full-fledged AI employees that automated large components of user research, complex data analytics, and even core strategic planning. In my last week, I had lunch with our COO and all we discussed was the latest agent skill he had built for summarizing customer feedback.

Eventually I started asking people outside the SF tech circle how they were building or deploying AI systems in their work / life / companies. Whilst there were some mavericks, the general response I got was that “it’s a good second Google / email writer. Everything else is a gimmick” or “AI isn’t relevant to our business. It’s too technical and we haven’t been able to crack it”

Akshay Kothari, founder of Notion, described the same gap yesterday on LinkedIn

There’s an insane amount of alpha in: people in SF leaving SF to see how people and companies around the world are using AI — hint: most are still using it primarily as search. And people outside SF spending even two weeks in SF — hint: many are already living in the future. The gap right now is staggering. The global diffusion of AI is a massive opportunity.

The irony is - AI was supposed to be the great equalizer.

With natural language, a structured mind, and clarity on what you want, capabilities that were constrained only to those with capital and exceptional talent (i.e. big American unicorn companies) are now genuinely accessible to all. I can personally vouch for this, having built a tech platform soup-to-nuts (with paying customers) in 45 days, whilst shifting countries, for less than $5K USD. And I’m not even a software engineer.

Here’s the bottom line - the longer we put off adopting AI and making it our company’s highest priority, the more revenue / customers / long term value we are leaving on the table. The gap between what AI native vs non native companies can do will only widen because these capabilities have a genuinely compounding effect.

So how do we start bridging the gap? And should we even care?

Let’s start with the second one. Here’s my prediction - companies that focus heavily on building AI, either internally or externally, over the next two years will accrete far more enterprise value than any of their competitors. They will become the de facto leaders in their industry because they will either offer better price, better quality or most likely, a combination of the two.

The people who support them in this endeavour will build careers that are lucrative in the future. This is the truest in traditional industries, where AI adoption remains embarrassingly low.

As for bridging the gap, the first and most decisive step you can take today is to learn what matters most from the human experts who are already at the other side, operating on the frontier of AI.

That’s why we built Luminary, where we connect people who want to build genuine AI competency with vetted operators from the best companies in the world. We train and educate leaders on the most relevant and cutting edge areas in AI. For instance, in our Agents for Superproductivity course, we teach you how to build reliable and effective internal agents that complete tasks in the real world (running marketing campaigns, creating product strategy, doing financial analysis, complex project management, etc.)

You compress 3 years’ worth of practical insights into 30 days.

And more people outside SF get access to what’s only available to a select few today.

That’s AI knowledge diffusion. And that’s what we are building towards.

Akhil