Head of Content
A leading thinker in the sphere of AI explains how many businesses will not see it merely as a means of cutting costs
In February, a ground-breaking experiment was undertaken in which 20 experienced lawyers were pitted against machines.
The task? To review five non-disclosure agreements, comprising of 11 pages and just over 3,000 legal clauses.
The lawyers – who were drawn from a range of major US firms including Goldman Sachs and Cisco - were given four hours. In fact they took an average 92 minutes, with the quickest completing the task in 52 minutes and the slowest taking well over two hours.
The lawyers achieved an 85% “accuracy” score, in terms of spotting risks or errors within the documents.
And the machines? They achieved an accuracy score of 94%.
But the killer point in favour of these artificial intelligence (AI) lawyers is that they took just 26 seconds to complete the task.
If an AI lawyer is 200 times quicker, significantly more accurate and costs next to nothing, surely the answer is to sack the humans?
“From a purely mathematical point of view the outlook for humans in the workforce is bleak,” says Dr Vivienne Ming, a San Francisco-based neuroscientist and entrepreneur, regarded as one of the world’s leading thinkers on the theme of machine learning.
“It takes 20 years to raise a human being. You can build a fully operational, self-improving AI function in 20 months.
“Why would I pay for lawyers, financial analysts or doctors when artificial intelligence can do their jobs better, far more quickly and for less?”
The ability of computers to scan, sift and sort vast quantities of data, and then draw meaningful conclusions from it, lies behind the many breakthroughs of Dr Ming. Her early work focused on facial recognition technologies, but data is at the core of her business successes, many of which are linked to health or improving welfare.
Her projects include AI systems which process huge quantities of blood sugar readings to predict dangers for diabetics. Another uses facial recognition software to analyse users’ moods: one aim is to help predict when bipolar sufferers are at risk of an episode. She has created tools for effective global recruitment – that rely in part upon trawling through reams of social media – and others that can scour tens of thousands of business websites to determine, for example, the gender split of the executives. She has also worked on a scheme aimed at helping orphan refugees connect with family members.
She has a clear vision of a not-too-distant future where people are even more intimately connected via technology and where “intelligent” functions – as well as mechanical – are performed by machines. And this will involve human roles becoming redundant.
“So far what we have seen is a process of substitution where machines do replace people. And, if you take a short-term view, that is understandable. It maximises productivity to get rid of the people and replace them with intelligent machines.
“But with an eye to the medium and longer term this substitution is wrong. It steals from the future of the business. The best businesses will see AI as augmenting, rather than substituting, a human workforce.
“AI is an amazing tool. When used properly, it will take human understanding and build on it. It will know about me and my behaviour and predict my mistakes.
“But you will still need humans in the loop. Humans bring emotion and social intelligence, and an ability to self-criticise and self-assess. Ultimately AI is not creative.”
Dr Ming’s vision of a world in which AI supports human activities and endeavours is clearly articulated in her many papers and speeches. She is an optimist. “The best technology makes us better people once we’ve started using it. Take mapping. I can get off a plane in Geneva and a map will take me to my meeting. It’s helping me achieve something, but it’s informing me too.”
Whether business will want to take the longer term view – rather than reap profit growth from a quick substitution of humans with machines – is moot.
“There’s a lot of cynicism within some large companies,” she acknowledges.
“But the companies that will turn out to be the Googles of the future will be the ones that figure out how to be inclusive.
“I’m not delusional. At the moment there is a small fraction of the world’s population which I call the ‘creative class’. They have the luxury of choice. Most of them live in western Europe and north America.
“In our unequal world, choice is not fairly distributed. Artificial intelligence could help reduce that inequality. The child who will grow up to cure a terrible disease might just have been born in a slum in Dakkar or a favela in Brazil. Artificial intelligence can help find that child.”
Dr Vivienne Ming is a neuroscientist, author, entrepreneur and founder of Socos Labs, a thinktank which focuses on machine learning and its impact on society. She sits on the boards of several companies. Dialogue interviewed Dr Ming at Gleneagles where she spoke at Entrepreneurial Scotland, an event sponsored by Cazenove Capital. To find out more about our events for entrepreneurs, please contact firstname.lastname@example.org or email@example.com
Head of Content
This article is issued by Cazenove Capital which is part of the Schroder Group and a trading name of Schroder & Co. Limited, 12 Moorgate, London, EC2R 6DA. Authorised by the Prudential Regulation Authority and regulated by the Financial Conduct Authority and the Prudential Regulation Authority. Nothing in this document should be deemed to constitute the provision of financial, investment or other professional advice in any way. Past performance is not a guide to future performance. The value of an investment and the income from it may go down as well as up and investors may not get back the amount originally invested. This document may include forward-looking statements that are based upon our current opinions, expectations and projections. We undertake no obligation to update or revise any forward-looking statements. Actual results could differ materially from those anticipated in the forward-looking statements. All data contained within this document is sourced from Cazenove Capital unless otherwise stated.