I’ve been seeing a lot of chatter right now about the lack of ROI from AI. I want to reframe that conversation, because what we’re experiencing is exactly what happens with every major technology wave.
Every technology develops in three phases. Phase one is the intellectual problem. This is the R&D phase, where the technology solves something that couldn’t be solved before. For AI, that was IBM Watson about 15 years ago. This phase always ends the same way: the technology works, but the infrastructure can’t support it at scale.
Phase two is the beauty and user experience moment. This is where we are right now. For AI, it started three years ago with ChatGPT. Suddenly anyone could access something that felt like magic. This phase always crashes too, and we’re already seeing the early innings of it with AI.
Phase three is the scale phase. This is where real monetization happens and the technology becomes essential infrastructure. For AI, our prediction is that it’s still about seven years away. That’s a long time, and it’s also an enormous opportunity.
The companies that win phase three are almost never the ones that dominated phase two. So if you’re frustrated by the lack of ROI right now, you’re in good company, but you’re also exactly where you should be. The question isn’t whether the ROI is coming, it’s whether you’ll be positioned to capture it when it does.
Join the conversation about this on my Linkedin, I'm tracking this closely and would love to compare notes.
Check out some of our recent content from top voices in the product space.
IBM CPO on Building Products for a World That's Five Years Away
The Conviction Bet No Customer Asked For and How Asana's CPO Builds the Case for One
Are you enjoying our Product newsletter? We are always striving to elevate our content and provide ample value. Please share your feedback, questions, and comments.
Good post and it tracks nearly perfectly with the transformational innovation cycles I’ve led for the past 30 years.
However, where I’d challenge it is in timing. 7 years is an eternity and while the pace of change management is yet to be overcome it is accelerating, too.
We built our company for the phase three from the ground up, and after two years have real traction at national-level orgs with phase three. I have been predicting that every major org by the end of 2027 will fundamentally either lease their business operations from a Frontier model provider or they will be implementing a stack like ours so they can own their business operations and advantage.