AI adoption in enterprise
Enterprises want real solutions to their business problems, not hyped incantations about AGI
Ethan Mollick recently shared his insights on LinkedIn about the conversations he has had with companies regarding AI adoption. His observations are telling and support claims I’ve made in other posts: the return on AI investment is likely to take longer than the hyperscalers, and Silicon Valley more generally, expect.
The narrative from AI accelerationists suggests that AI will transform everything, everywhere, all at once—a notion more suitable for a Hollywood tagline than for practical enterprise adoption. The reality is that while companies are interested in AI, they are still figuring out how to integrate it into their existing workflows, and how to restructure these workflows to fully leverage AI’s potential.
There are a couple important reasons for this caution:
Organizational Biases: Resistance to new technologies is common within established enterprises.
Concerns: Security, legal, and ethical consideration regarding AI technology need thorough examination.
These challenges are solvable, but organizational decision-making tends to lag behind technological advancements. Many AI proponents mistakenly equate rapid technological progress with swift enterprise adoption. However, improvements in technology do not necessarily correlate with immediate adoption.
Recent research from Bain underscores both Mollick’s claims and my observations. Bain’s quarterly survey indicates that while companies across industries have high expectations for the value generative AI can add, they are still in the exploratory phase. Executives are investing in talent and resources, but many companies are stuck in the learning phase, hesitant to commit significant capital without a clear return on investment.
The Bain report reveals a key insight: companies recognize AI’s potentnial, but struggle with practical application. This distinction is often overlooked by AI accelerationists. Understanding the importance of a technology differs significantly from knowing hot to apply it effectively. To overcome these hurdles, those working in enterprise AI must build close relationships with prospective customers and deeply understand their problems. This ensures that products that are being built address real needs. Vague promises about AGI transforming everything will not drive enterprise adoption. Instead, practical, customer-focused AI solutions will move the needle.
Your observation that “those working in enterprise AI must build close relationships with prospective customers and deeply understand their problems. This ensures that products that are being built address real needs” should be put on a plaque at the entrance to every company trying to deploy AI. AI should be subjected to the same analysis as any software - what is its purpose? and does it answer the need better than other options? Believing in the hype and being worried about being left behind are natural concerns - but they should be balanced with some analysis.