As incumbents race to add “AI features” to their applications, these costs are becoming more front and center. Large companies deploying AI features — Notion, Dropbox, Adobe, and many more have launched new products powered by AI advancements. Even larger companies like Google and Microsoft plan to reorient core product lines around these models (e.g., Google Search, Microsoft Office).
But every feature has a cost. The obvious costs are time and money. But less obvious costs like screen real estate, brand perception, and organizational scar tissue are often the most costly.
The screen real estate problem is illustrated with Google’s knowledge graph panels. There’s an extremely limited (and lucrative) space right below the search bar. It can be filled with ads, organic results, or something else (e.g., an AI-generated answer). You can run experiments (and Google does), but choices still have consequences. Notion’s AI features are exciting (at least for me), but how will power users or enterprise customers feel over time?
Another cost is organizational scar tissue — one unsuccessful attempt at a market (e.g., too early, wrong people) prevents a company from rationally entering the space again in the future. Take Google Code — a code hosting platform like GitHub that existed from 2006-2016. It failed for various reasons, but it would be hard for Google to launch a similar project today (even if the space is ripe for disruption). Being burned today on AI might prevent companies from utilizing it when the use cases become more solidified.
First-mover advantage is overrated, and deploying AI features too quickly (even though it might be straightforward) might haunt these companies. In a self-fulfilling prophecy, haphazard changes to longstanding core products might create the opening for incumbents to get disrupted.