What Is So Difficult About Public Goods?

They can't all be as ignorant as Mazzucato, can they?

That some are indeed as ill-informed as Mazzo is obvious - she does gain an appreciative audience after all. But that the wider policy-making sphere gets these things so wrong is not a grand advertisement for our education system.

So, the inscrutable Chinee manages to make an AI that works - so I am told, slowly, but it does work - on a Raspberry Pi and everyone falls about in horror. When, of course, the correct reaction is to have the massed marching bands out in the streets.

Yes, a trillion is knocked off the stock markets, $600 billion off the AI chip maker, and this is glorious. For this means that AI is cheaper than we thought it was going to be. The capitalists are going to be able to pocket less of the benefit than we thought. More of that lovely AI goodiness will flow through to us, the consumer.

Now, if the heathen Chinee were the only person who could possibly achieve this feat we might not be so wholly and awesomely satisfied. But that’s not, not in the slightest, how technical advance works. Once we know it’s possible then we know that it’s possible. That’s always 90% of the difficulty. Once we know it’s possible then working out how to do what is possible is not necessarily easy but it’s easier. We have an end goal, we’re not trying to explore the entire possibility space.

Which brings us to this:

China is building a gigantic laser-ignited fusion power laboratory that is 50pc larger than its US counterpart as the two superpowers spar for energy supremacy.

Basic research simply isn’t a competition - it’s a public good. Which is what brings us to Mazzo, of course.

A public good is something non-rivalrous and non-excludable. Something that once it exists cannot run out and also we cannot stop someone from using. As such it’s damn difficult to profit from. Thus capitalists won’t invest - or perhaps just not as much as we’d like them to - in the production of public goods. So, use government! Not necessarily to do it but perhaps to twist the system to encourage - copyright, patents, say. Or, yes, maybe pay labs to do the work.

Which is fine, obviously. Good arguments too.

But note what the argument actually is. We can’t make money out of public goods production. So, it doesn’t matter who produces public goods. It only matters that public goods are produced.

Fusion. We don’t know whether it can be made to work. Well, we do, but without having a star to power it we’re not sure. So, China spends a gazillion on proving that it is possible. Or not, as the case may be. We now all know the answer without having had to spend a gazillion.

Public goods production, especially in basic science, is not a competition, it’s a cooperation. It’s not just that we give not the proverbial shit if China does it we’re gobsmacked with delight at the idea that they spend the money and do the work to our benefit. For knowing that fusion works will be very valuable. Just the knowledge that it works will have that value. For once we know it does then replicating it - like some AI on a $30 chip - will be easy enough. It’s the knowledge that it is possible which is of the value.

The Mazzo Disease is to think that public goods are good - they are - and then to insist that both we must produce our own and also that there’s vast profit to be made by doing so. When the entire point of, the very nature of, public goods means we don’t give a toss who produces them because no one can profit from having done so.

Well, OK, 8 billion people profit from someone having done so but the people who did so don’t.

It’s astonishing how difficult some people find this. Including far too many professors.