Adam Smith Institute

View Original

Why not waste another £400 million on AI then?

Image Credit: https://claudeai.uk/

It’s entirely true that we have government because sometimes it is useful that we have government. There are things that can only be done collectively and with that element of compulsion that government power provides.

We are not, after all, anarcho-capitalists.

It is also true that there are things that government can do which we do not wish government to do. Like do those things that can - and will be - done better, faster, bigger, by other mechanisms than government. At which point let’s save the oppressed taxpayers of the nation £400 million, shall we?

Rishi Sunak is poised to increase taxpayer spending on artificial intelligence chips and supercomputers to £400m as he strives to boost Britain’s technology credentials.

The Government is quadrupling a planned investment package to secure equipment for a national “AI Research Resource”, The Telegraph understands, which will also be used to launch a new facility in Cambridge and improve a site in Bristol.

The increased investment comes as the Prime Minister seeks to present Britain as a global hub for AI safety to world leaders at a summit at Bletchley Park on Wednesday and Thursday.

This is not investment this is wasting money.

From another source:

Google, Amazon, and Microsoft funnel billions into rival AI companies. There is just so much money flying in this colossal AI arms race.

• Microsoft has poured as much as $13 billion into OpenAI.

• Amazon recently reported a potentially $4 billion investment in the Google-backed Anthropic.

• Now The Wall Street Journal reports that Google will slide Anthropic a cool $2 billion over time on top of its previous investments in the company.

The government money - our money, that lifted from our wallets - will arrive in two to three years’ time, be miniscule in market size, shrunk by the inevitable multi-layers of bureaucracy and, in all probability, be spent upon the wrong things. Late, trivial and worthless.

There is indeed an argument - Professor Mazzucato does at least start from a logically firm position - that government should spend on useful things that market participants do not. Those public goods which are undersupplied by private economic actors. If we’ve $21 billion in just this week’s news story about AI then adding £400 million of our capital just isn’t relevant now, is it? It’s just an imposition on our wallets.

But more than this. AI is an investment frenzy right now. It’s possible to float a company - we really do see this all the time - where the addition of the word “AI” to the prospectus gains access to market capital. The AI company of great profit but no one to know what it is would raise money right now. Actually, someone’s already done that and raised $600 million.

Let’s not use government to do what the market is already doing bigger, better and faster. Horses for courses and all that, reserve governmental powers of compulsion and inefficiency for when there’s no other viable option.