It's no secret that document processing is a crowded space. If you look at AI and deep learning more broadly it feels like every major cloud provider is making some kind of play. VCs and founders alike often ask me how I can stay in business when there are these massive players trudging around in my space.
I often remember a call I had with an IBM executive many years ago. Embarrassingly, I didn't remember who he was when I took the call, so when he asked for feedback on IBM's AI strategy, I answered candidly. That's to say that I didn't think it was very good. In my mind, IBM made a fundamentally bad bet against deep learning for the power of marketing. Simply by looking at the ratio of their marketing spend to their R&D spend, sharp minds could sniff out the problem.
His response was brusque. "IBM has already won. Small startups like you should shut up or get on the boat."
At the time it worried me, and I thought that billions in marketing might just be able to overcome a fundamental mistake in technical strategy. Today? I realize that every top AI organization realizes that the mission is too big for anyone to tackle alone.
Hear more about it, with an easter egg about my embarrassing encounters with Spike Lee and Mark Cuban.