AI will rule the world?
Posted on 29th May 2018
In the 16thcentury Spain created the world’s first global empire. Then, in the 19thand early 20thcenturies Britain dominated the world. Since then it has been the turn of the Americans and as the 21stcentury evolves it’s going to be the Chinese and the Indians. Underpinning this dominance, in every era, was economic strength, and it’s clear that the economic powerhouses of the future will be dependent on – if not masters of – technology, and in particular AI. To demonstrate this, consider the following stats, culled from a variety of sources.
First, let’s look at the impact of AI on growth in the various parts of the world between now and 2030. PWC estimates that AI will increase growth as follows:
Latin America - 5.4% - equivalent to $0.5trn.
Rest of world – 5.6% - $1.2trn
Northern Europe – 9.9% - $1.8trn
Developed Asia – 10.4% - $0.9trn
Southern Europe – 11.5% - $0.7trn
North America – 14.5% - $3.7trn
China – 26.1% - $7trn
Next are some figures from Capgemini, considering the benefits accruing from AI. They report that companies are seeing a benefit of greater than 10% in the following areas:
Increased operational efficiency
Improved employee productivity
Reduced operational costs
Greater legal and regulatory compliance at lower cost
Reduced false positives
Thirdly, Accenture has estimated the impact of AI on industry growth by sector to 2035, assuming a “steady state adoption/implementation scenario.” Unsurprisingly, “information and communication” shows the biggest growth, with “education” at the other end of the scale.
Education – 1.6% annual growth to 2035
Arts, entertainment and recreation – 3.1%
Utilities – 3.1%
Accommodation and food services – 3.2%
Agriculture, forestry and fishing – 3.4%
Construction – 3.4%
Healthcare – 3.4%
Professional services – 3.8%
Wholesale and retail – 4%
Financial services – 4.3%
Manufacturing – 4.4%
Information and communication – 4.8%
Finally, while this growth is generally to be welcomed, it’s worth considering the impact of AI beyond its obvious implications for the economy. Here are just a few projects where AI/machine learning has helped identify or solve problems in various parts of the world.
In India, algorithms on tax data allow the tax authorities to identify firms possibly indulging in unlawful activities to try to minimise their tax liabilities, thus leading to physical audits to check whether this is actually the case.
Machine learning is being used to predict food security in Malawi.
Algorithms can be used to predict credit worthiness better than analysis of an individual’s current credit history. This can help to reduce gender gaps in access to credit (women are statistically less likely than men to have a credit history).
In the same area (credit worthiness), a machine learning algorithm used data on 5,500 indicators of mobile phone usage to predict the likelihood of loan repayment and thus extend credit to those who don’t have bank accounts and hence no credit history.
Again using mobile phone data, models have been created that predict how and when people evacuate during disasters.
There are hundreds more of these, but, of course, there is just one thing that nags away at the average man or woman in the street (i.e. me!): in amongst all this academic and government research and analysis, how safe is AI going to be for me – can I trust big business and government to behave ethically? And that’s a subject for my next blog in a few days’ time.
Nikola Kelly, MD, Be-IT
Posted in News
.. Back to Blog