As we're witnessing in real-time, Microsoft and other GAFAMs are plucking every remotely interesting AI startup remotely interesting out of the market (either by acquiring them or having enough investment to control them); open innovation seems the best possible way of catching up with the fast-paced market turnaround.

The concept of open innovation is nothing new. Popularized by Henry Chesbrough at UC Berkeley, the core idea is that strategically bringing in an innovation built outside your company (whether by acquisition, investing in a controlling stake, or licensing the IP) is often decisively faster and safer than trying to innovate in-house.

The success of the idea of open innovation comes from the belief that 90% of innovation risks can be mitigated by timing strategic purchases outside of your company. On paper, it also solves most of the innovator's dilemma. Not only did Microsoft "acquire" Open AI –I know it's not strictly an acquisition, but for all practical matters, Microsoft now fully controls this company– but they also invested in half a dozen of Open AI's competitors just to fully hedge their bet.

So what's wrong with all this?

Not much, actually. Open innovation does really work quite well. (There's a 'but' coming...) But it's only one tool in a vast toolbox. And the problem, as always, is that when something seems to be working quite well, corporations just want to go full in and only do that. While in the context of AI, open innovation is certainly one of the best approaches for laggard companies that cannot afford to build LLMs on their own, fine-tune them, and bring them to the next innovation cycle, it's not going to solve other types of problems.

Open innovation is the right tool when a technology that was supposed to need ten more years to mature and solve obvious market problems is suddenly pushed into the market by an unexpected breakthrough.

The open innovation zone is even fairly limited if you think about it in terms of a strategic innovation portfolio zoning:

The open innovation field strictly overlaps with the technological "New Frontiers" field.

Corporations and innovation departments always need to deeply understand this: innovation is a vast field with no "best solution" or unique "winning strategy," only ad-hoc answers and tools for solving specific problems.

The link has been copied!