Justification for EU laws and policies often comes from concerns about fragmentation if Member States legislate in ways that fragment business markets or have widely different levels of social protection. Yet the EU's motto, unity through diversity, shows the tension between these two attributes. Should the EU be trying to promote lowest common denominator legislative compromises or rather wait for successful approaches to emerge from amongst Member States who will develop good and bad legislation or even no legislation? This is the familiar dilemma between ex ante and ex post, regulating before an area is mature and trying to regulate afterwards when many national laws are in place across a wide spectrum of positions and effectiveness.

What is the balance between harmonising regulation, and harmonised approaches to investments and political priorities? Maybe the degree of harmonisation needed differs across all those. Or do we need consistency?

The rise of the professional politician incentivises legislation and an activist posture. How far can policy affect the Darwinian processes innovation, where there are by definition winners and losers? As The Economist noted several years ago, state intervention to foster the then fashionable notion of clusters has not been very successful, suggesting that the orthodoxy prevailing since the 1970s of not "picking winners" still has value. Governments copy each other, but ultimately innovation is not very "fair," I read recently.

On the other hand, there is the old anecdote told about a minister in charge of Japan's legendarily interventionist MITI ministry in the 1980s. Reproached for "picking winners," he smiled and said that his ministry never picked winners: "On the contrary, we set up races between different companies and the winner of the race becomes our preferred choice". Prof Mariana Mazzucato's recent work around the theme of the entrepreurial state argues for a more MITI-like approach by the state.

This debate was raging in 1984 when I first worked at the Commission in the era of "national champions," "strategic targeting" and MITI's then frightening - to policy makers -  but I think unsuccessful "5th Generation Computing Programme."  Thirty years later I don't see Japan as being very visible in what we would now call artificial intelligence.