One of the hottest topics in business these days is artificial intelligence. One can hardly pick up a business publication of any sort and not trip over an algorithm. AI is being used to do everything from writing articles to running chatbots to protecting against fraud. There is one problem with AI, though, and that’s our topic today.
You’ve probably encountered something that’s the product of AI. A fair number of game summaries one finds in the sports pages (physical or digital) are, in fact, written by machines. Same with many company summaries in the financial section. The main problem with these pieces is that they’re great at populating a template with all the facts and not so great at figuring out the “why.” You might also have used an online chat function to get some customer service support. More often then not, that’s AI at work as well. But that’s not the business problem I want to discuss.
The problem with most of the AI solutions I read about is that they’re all geared toward helping a business but they’re not focused at all on helping the customer. If you’ve ever wandered into an AI-driven customer support phone line you know what I mean. Get outside of what the algorithm can handle and your blood pressure is sure to soar. While the bot on the other end knows all about you if you’re able to identify yourself in the way the AI is designed (frequent shopper number, etc.), if you don’t know what phone number was used to create the account or you’re a frequent shopper without a frequent shopper ID (some folks don’t live being tracked, you know), it’s hard to get support. Humans are still better at solving many non-standard requests.
I get that sharing all your data – what you read, what you watch, where you go, what you eat, etc. – can help a company give you better recommendations. The problem is that many of the companies use that as a pretext to sell you products you might not really need. Can any of us really know how the data was used to create a recommendation? When a fitness app tells us we’re having sleep issues because our data says so and says we need to buy a new mattress, can we trust that or is it an affiliate deal that brings the fitness app a commission? Maybe we just ought not to have that nightcap instead if we want to sleep better?
I think the use of AI in some areas is fantastic. Fraud protection, for example. It’s easy for AI to spot something that’s out of place in your credit card use and send you an alert. That’s customer-centric. Using a bot to cut costs while providing a lesser experience isn’t and that’s my issue with much of the AI work that’s going on now. What’s your take?
Welcome my son, welcome to the machine.
Where have you been? It’s alright we know where you’ve been.
You might think from Roger Waters‘ lyrics he was thinking about privacy. His take was on the recording industry but the song came to mind as I was thinking about robots. I’m hoping today to get you to ask yourself about them as well. Specifically, are you better than one, or at least the one that can replace you in the business world.
Before you laugh that off, consider a few things. We have cars that are driven by robots and they will absolutely do a better job than a human driver most of the time. Sure, there will be times when a road is temporarily closed and the mapping software will be lost. The car will need some immediate human input but that’s probably the exception. Most trips will be safer and less stressful. Bots don’t get tired or distracted. Can you say the same?
Bots can write and do so in a way that is pretty indistinguishable from humans writing the same stuff. Around 10% of Wikipedia articles were written BY ONE BOT. How many people do you know that are that prolific? You’ve read sports stories both on the web and in newspapers that were written by bots interpreting statistics to produce a game narrative. Financial reporting is being done by bots too. They do an excellent job of sticking to the facts, uninfluenced by some PR flack’s spin. I’ll admit that I get weekly reports on Google Analytics that are written by a robot. They’re not great but they do a good job of calling my attention to things that require further investigation. I imagine they’ll continue to improve. Much media selling is now done by robots and programmatic ad buying is forecast to have robust growth as more publishers, advertisers and agencies embrace programmatic technologies. As a result, Magna Global projected that the volume of transactions will grow to $17B by 2017 in the US, of which $10.5B will be RTB-based. Those are lots of sales, sales support, and media buying jobs going away.
Am I bringing this up to depress you? Not at all. We need to think about how we can do many things – still – that can’t be replace by a machine. Nested phone menus for customer service are a form of automated response – do you know anyone who prefers that over talking to a human for service? Bots are only as good as the algorithm that drives them. While we’re heading to a world of artificial intelligence and algorithms that will self-improve, I suspect that we’re a long way off. That said, we need to emphasize the human parts of every interaction. A bot could present the facts of this post but I don’t think it can provide tone and nuance (or ramble quite as much).
So that’s the question for each of us, isn’t it? Can we be replaced by a bot or are we adding things that are uniquely human to our business – and non-business – lives? What’s your answer?
Sometimes I’m convinced that the most successful businesses have no idea what they’re doing. Oh sure, if you asked an executive about their strategy, he or she would probably give you whatever is in their planning document verbatim, but I think that’s crap. I think they’re telling you what they believe to be the truth but in fact may only be accurate in their minds for that moment. Instead, I think the most successful companies are masters of dealing with the utter chaos of the business world and not being too anchored to any one detail of a plan. Yes, planning is important, but so is reality, and that often means dealing with something for which we had no plan. So why bother planning? Continue reading