Economic Impact Of AI

Economic Impact Of AI
Implications for markets, the economy and people

We will now shift focus and talk about what AI means for existing market structures, the macroeconomy and for ordinary people.

The four Podcasts in this module will look at the following four topics:

  • The AI value proposition: The product concept that lies at the heart of AI is the commoditisation of above-human intelligence. It is important to distinguish this from the dumb hardware infrastructure elements – computers, robots and physical machines – that act as delivery conduits.To understand the profound nature of this value proposition, imagine one day in the future going to a recruitment website to hire a legal expert to investigate whether you have a case against a supplier who refuses to pay. Now imagine that the ‘expert’ is not a human lawyer, but is a dedicated instance of an artificially intelligent system created specially for you. You pay a  fixed fee for expect counsel which gets smarter the more it is used.
  • Market-level effects: How many industrial sectors will be impacted by AI? Is there a way to escape? What is the best way to take advantage of AI? Can we trust the free market to protect itself? Will the actions of good actors outweigh those of the bad?
  • Overall macroeconomic considerations: If we cannot even measure the digital economy then how can we say we understand how AI will affect the global economy? Are we in a situation where “we’re going, but we don’t really know where we’re going”.In this Podcast we will reveal just how fragile our grip on the digital economy really is and how systemic measurement problems might prevent us from properly understanding what is happening.
  • Policy factors: Here we will talk about how the mass deployment of AI technology could affect social cohesion, precipitate job displacement and create a ‘jobless generation’.We will also discuss the Universal Basic Income (UBI) and whether we are approaching the point where policy makers need to rethink the concept of work and the purpose of growth.

There seem to be two opposing views emerging regarding how AI will impact the economy:

Some people – like Elon Musk and Stephen Hawking – are warning that AI technology could develop into something that if used incorrectly, could prove disastrous not just for the economy, but for humankind.

The opposing camp consists of people like Mark Zuckerberg and Ray Kurzweil who do not quite understand why their peers are so worried: their position is that while there are some factors that will need to be managed, AI is without question going to deliver tremendous benefits to the economy and dramatically improve the human condition.

Silicon Valley gets some things wrong

That we are seeing a sharp difference opinion about how AI will impact the economy at such an early stage in the market’s development is interesting because we did not see that in 1995-1999 when the Internet first got going and we did not see it in 2007-2012 when social media took off, either.

But, as we now know, there was good cause to be suspicious about the claims made by Silicon Valley VCs and startups before the dot-com bubble burst.

And there was in hindsight good cause to be concerned about how the mass adoption of social media would affect the minds of millions of users: Sean Parker, a co-founder of Facebook and Roger MacNamee, who provided seed funding for Google and Facebook are now openly asking ‘What have we done?’

So maybe we should think critically about the warnings vocalised by people like Elon Musk, Nick Bostrom, Martin Rees, Bill Gates and Stephen Hawking are saying?

The purpose of this module is to unpack the arguments on both sides and then view them in the correct micro and macroeconmic context.

We are going to also look at how to measure the impact of AI, which is more complex than you might think:

We have a measurement problem: GDP cannot accurately measure the impact of digital markets

Here’s something to get your mind working: around 500 million people visit Wikipedia each month. And 50,000 people work hard every day creating and editing content.

There can be no doubt that Wikipedia – as an entity – delivers real value to real people and employs real people who are using useful, valued work.

So you’d probably assume that Wikipedia’s contribution to the economy is part of the GDP calculation – which is how policy makers measure the economy and manage economic growth.

Well, you’d be wrong: the contribution that Wikipedia makes to the global economy is zero.

Remember that the site is free to use and carries no ads which means that no spending – either by consumers or businesses – can be directly attributed to it.

The donations made to keep the site running are not included in GDP figures.

So, quite literally, Wikipedia is a zero value-added service.

This is clearly absurd,  but it is just one of many examples.

Here’s another: total retail spending on music has about halved since 1999 but the real value delivered to music consumers has increased dramatically: services like Spotify, Pandora and Apple Music collectively offer a far superior quality of service than that available in 1999. As far as the vast majority of consumers are concerned, digital is simply better than physical.

But because GDP measures spending – and spending has gone down – then this means that we have no way of measuring the increased value that the music industry has delivered.

In fact, if we use GDP as a metric to assess the real societal value delivered by the music industry then we conclude that value delivery has decreased – when it is obvious that it has increased.

This does not make any sense.

The uncomfortable reality is that our deeply-entrenched way of measuring the economy is dysfunctional when applied to digital markets.

Economists might say that the impact of the Internet is small – but that is in many cases because most economists are using metrics – like GDP – that cannot measure what is really happening.

This problem will become a lot worse when AI begins to be deployed more widely – the displacement of labour and disruption to markets may remain invisible to economists for years before second or third-order effects start showing up in other metrics.

How AI will impact people

Perhaps most important of all, in the fourth and final podcast of this module we will talk about how AI might impact ordinary people.

We will look at the idea of a Universal Basic Income but we’ll also explore some new thinking on what people might do if we ever get to a point where the parts of the economy that are needed to support day-to-day life have been mostly automated and can be operated at very low cost.