AI doesn’t replace YOU
On April 6 2018 Jacqueline Rayner posted this tweet
It has been retweeted by thousands and is quoted all over the internet. The most striking thing about this post is that without any context, background or follow-up, almost everyone who reads it will know exactly what she is referring to: Amazon’s “smart” algorithm which makes sales suggestions to each individual customer based on their previous purchases, items viewed, current trends and so on. This is AI in action.
Amazon will tell you that this AI engine is very effective, leading to a relatively high number of further sales. Much of the time, the algorithm will hit the mark and make sensible, useful suggestions which may be something the customer actually needs. And then there are times like this, where the algorithm does its stuff, and works exactly as it’s supposed to, but the result is not only useless but ridiculous. But to recognize this as ridiculous requires judgement which is a human quality. Computers are great at number-crunching and rule-following, and given time they can learn intricate behavioural patterns so that they can mimic judgement, but at the end of the day, they remain machines.
Peter Thiel, one of the founders of PayPal and Palantir Technologies, elaborates on this in his book “Zero to One”:
To understand the scale of this variance, consider another of Google’s computer-for-human substitution projects. In 2012, one of their supercomputers made headlines when, after scanning 10 million thumbnails of YouTube videos, it learned to identify a cat with 75% accuracy. That seems impressive — until you remember that an average four-year-old can do it flawlessly. When a cheap laptop beats the smartest mathematicians at some tasks but even a supercomputer with 16,000 CPUs can’t beat a child at others, you can tell that humans and computers are not just more or less powerful than each other — they’re categorically different.
To paraphrase Thiel, there are many tasks which can be done by computers using AI which are time-consuming or monotonous, or simply only made possible because of machine power. Handing these tasks over to the computer systems frees us up to focus our resources on judgement-based tasks, those which even the most powerful AI machine is not able to fulfil.
It is easy to laugh about the ways in which AI misfires, but we are able to do so because it has become such a pervasive part of our lives that we take it for granted. There are important reasons as to why companies and governments invested huge resources into the development of AI and we are only just beginning to reap the rewards. One such example is the way in which apiculture specialists Bee Futures together with AI developer Amesto NextBridge have made use of AI technologies to help overcome the problem of colony collapse disorder, which has made news headlines since the early 2000s. Globally, we are highly dependent on bees for crop pollination. This means that when bees are in trouble, so are we! Bee Futures discovered that it is possible to interpret bee behavior by studying their individual movement patterns. AI made it possible to translate this research into useable data: by placing micro cameras in and around the hive, and connecting them to computers, their AI engine has the capacity to analyse the movements of each separate bee and bring the results together to present an overall pattern which can then guide the farmer as to the best location to place or reposition the hive, as well as offering guidelines as to the necessary biodiversity to sustain the colony. This would simply not be possible using human power alone, there are too many data points to collect and analyse, and too short a time-frame in which to do so.
Sylvain Duranton, global leader of BCG GAMMA, a unit dedicated to applying data science and advanced analytics to business, emphasised the importance of humans and AI working together to bring about the best possible outcomes, in his TED speech back in 2019. He quotes MIT research that states that the large companies investing in AI are doing so not simply to save money, but to improve their business through greater efficiency and growth, which is best achieved by creating a cooperation between man and machine. This leads to the question, why isn’t all AI being built in this way? Duranton reminds us “that constructing such a system can be long, costly and difficult…BUT the reward is huge”.
At Okapi we are embracing Duranton’s philosophy. While we believe AI – including our own – is the future, we know it shouldn’t be a supplement for human knowledge and experience. Just as Bee Futures cannot possibly follow so many individual bees and all their behaviors without the assistance of AI, you cannot be expected to follow all the movements of all your tenant “bees” outside of the “hive” of your asset, and there are so many factors to be taken into consideration when calculating CRE risk which is where the power of AI comes into its own. You can learn more about the effectiveness of a collaborative effort in the post “Our tag-line as a philosophy for our business”. When working in tandem, our innovative technology’s groundbreaking risk assessment and prediction capabilities enhance your professional knowledge of your tenants to achieve the best results.
Our sales team can set you up with a free trial to help you discover what the pattern of your “bees” means for your “hive”, with a full risk-assessment report for up to 10 tenants, so contact them today at [email protected] and discover how our AI can innovate your business.
(Image credit: Hitesh Choudhary for Unsplash)