As referenced in my last piece, the ultimate winners in the Western were not the gunslingers themselves, but the farmers they had come to save. Like farmers who are always with us, the analogy extends to those businesses for whom change is more gradual but potentially longer lasting.
The first thing to note is that AI as a concept in the workplace is not new; witness the deployment of robotic technology to replace people in repetitive work environments, such as a factory assembly line, which has been going on for a generation. What is different with the new iteration is the capability of a machine to not only perform a task continuously (no coffee break needed), but also to make some simple deductions based on the experience it gains and improves its performance as a result.
One of the areas that has been touted as a likely winner is in drug discovery. The idea is that pre-clinical studies of any new compound or molecule will be able to utilise this new ‘intelligence’ (which is really just extra compute power) to run a multitude of simulations more quickly and therefore more cheaply. It remains to be seen if this leads to more breakthrough therapies and treatments, but certainly everyone’s R&D budget will stretch further. This is hard to handicap from an investment point of view, because it is a bit like a football club being able to buy loads more players – as a Chelsea supporter, I can attest to this not necessarily leading to a better team!
The most striking difference that we are all likely to see very soon, if you have not already, is around Chatbots. That frustrating virtual ‘assistant’ – a complete misnomer in many cases – that has replaced a person is about to get a lot smarter. While I am no fan of the ‘Buy-Now-Pay-Later’ (BNPL) model of companies like Klarna, what they have achieved with their AI assistant in just one month is phenomenal. It has handled 2.3m conversations, that’s two-thirds of their total, doing the job of 700 full time call centre staff and customer satisfaction is on par with human agents. It is more accurate at “errand resolution” (their words) leading to a 25% drop in the total number of inquiries. Resolution of a call averages 2 machine minutes, versus 11 human equivalents. It is already available in 33 languages and of course works 24/7. Imagine being able to deal with your Bank, Insurance or Utility company quickly, easily, and accurately – soon to be a reality.
The entire financial services sector will be a huge beneficiary in terms of ledger transfer, trade settlements and the myriad other administrative functions that they are required to perform, although the security requirements of certain data will prevent some elements moving as fast as they might like.
If those first two examples were about efficiency and cost savings, another sector might use AI to expand its sales and revenue as well. And that is retail. The bigger you are, the more opportunity you possess. Two months ago, I wrote that one of my picks of the year is Walmart; the world’s largest retailer. Here is just one way they will use AI. All stores already carry cameras that watch what we do while browsing along the aisles. But now they are going to be watching the product on the shelves as well. Once something runs out, it will report back with news that an item is now ‘out of stock’ allowing a human to refill it at that point, rather than waiting for their pre-determined restock timetable. A recent study across all types of retailers estimated that they have an average of 8-14% of their inventory not on the shelf at the time a customer might want to buy it. That’s worth over $1tr annually. With AI systems deployed alongside the cameras, store managers will soon be able to know for every single item, not only what is being bought, but any patterns of purchase that occur in the time of day it is bought, if Tuesdays sell more than Thursdays. The manager will know if a particular hair product is most popular among red headed women in their 30s on Friday lunchtime.
Retail stores will finally be able to match the type of knowledge that has previously been available to their online competitors, who have been able to track your digital footprint for years and then target advertising or adjust price and selection accordingly.