AI: Startup Vs Incumbent Value

AI: Startup Vs Incumbent Value

In the ever-evolving landscape of technology, the distribution of value often shifts between startups and incumbents. Over the years, we have witnessed this shift in various waves, from the first internet wave to the rise of mobile and now the emergence of AI. Surprisingly, the prior wave of value from AI predominantly went to incumbents, despite the significant startup activity in the field.

When we look back at the first internet wave, it was the startups like Google, Amazon, Paypal, and Facebook that captured most of the value. However, incumbents such as Microsoft, Apple, IBM, and Oracle also managed to extend their franchises onto the internet and secure a portion of the value. The split between startups and incumbents in this wave was roughly 60:40 or 70:30 in favor of startups.

In the mobile wave, the dynamics shifted, and most of the value went to incumbents like Apple and Google. Startups like WhatsApp, Uber, and Instagram did manage to capture a significant portion of the value, but the split was more skewed towards incumbents, with a ratio of around 20:80 in favor of incumbents.

Crypto, on the other hand, witnessed a 100% capture of value by startups. Bitcoin, Ethereum, Coinbase, and other cryptocurrency companies emerged as the major players, while existing financial services and infrastructure companies had limited participation in value creation. This wave showcased the potential for startups to dominate an emerging technology space.

When it comes to AI, the trend seems to lean towards incumbents once again. While there were many "AI first" startups in the last decade, the truly transformative AI applications landed with companies like Google, Facebook, TikTok, Netflix, and Amazon. These incumbents leveraged their existing resources and customer base to create AI-powered products and services that dominated the market.

To beat an incumbent as a startup in the AI space, you usually need to build something dramatically better or focus on a brand new customer segment or distribution moat that the incumbent cannot serve. In other words, you need a 10X better product. Incumbents often have the advantage of distribution, capital, and pre-existing product moats, making it challenging for startups to compete.

However, the landscape may be changing. One possible reason why incumbents have been successful in the AI space is their data advantage. As companies now have access to the broader internet as an initial training set and are switching to models that work well with smaller data sets, the data advantage of incumbents may diminish.