Entrepreneurship Through Acquisition in the AI Era
Like previous technologies, AI adds immense value to legacy businesses when deployed effectively. However, AI has proven extremely difficult to integrate with many organizations’ current infrastructure, which is partially the reason why many believe AI-native businesses can outcompete incumbents in the same market. This is often a suboptimal approach. AI is far from being considered reliable enough for must-be-right workflows. Consider a model that gets 95% of a machine shop’s estimate right. If it fails to account for, say, six hours of setup, the shop can win the order precisely because it has quoted a price at which it will lose money.
Of course, as the AI frontier develops, AI’s reliability will increase. Even if AI generally becomes more reliable than the average professional, its intelligence is worth little without the data that informs how to do good work in a domain. That data is owned by incumbent businesses. AI-native startups have found it difficult to compete without this data, even if their offering is faster and cheaper. Ergo, they have an incentive to buy access to that data, and some are offering millions to do so. Crosby, an AI-native law firm, bid for assets of Weil Gotshal that included internal data, and separately sought to hire its lawyers. That hiring push reflects that owning the data alone often isn’t sufficient to develop a competitive product. Much of the value lies in the hard-won judgment behind those records. A file can show that an order lost money, but the estimator who priced it can explain what she missed and how she would quote it today.
Continue reading ....