Jeff Bezos spoke this week at Recode’s Code Conference in California about the “gigantic” potential of artificial intelligence. Pressed by moderator Walt Mossberg to speculate on where this technology is headed and what it means, Bezos went so far as to say that we may be on the “edge of a golden era”. Were it not real life I’d have sworn I was watching the opening scene of a sci-fi horror film, complete with scary technology (AI), calming platitudes (“humans are much more efficient”) and sinister villain (guess who).
Contrary to the titillating trailer, however, I think the future of artificial intelligence, at least as far as supply chain strategists are concerned, is bright, and that Bezos is our friend.
AI and customer demand
Amazon reportedly employs 1,000 people in artificial intelligence. Some are presumably working on the company’s Echo wireless speaker that offers consumers an alpha version of the robot made famous by Will Robinson on TV’s classic Lost in Space. The device listens to you and speaks back. It does things like turn off lights, report on traffic and order things (from Amazon, of course). Most reviews seem to think it’s pretty cool and the company claims it’s selling fast.
If the Echo was only capable of voice recognition and remote electronic transactions it would be a mere novelty or, at best, a worthy convenience. Backed up by artificial intelligence, however, it could be something far bigger.
Bezos’s comments on the promise of AI specified three technology foundations that make it all possible:
- Compute power
- Training data
The first has been around since the time of Euclid and the second has been working for 50 years thanks to Moore’s Law. The third sounds innocuous enough, but is in fact the crown jewel of Amazon’s eternal strategy, which is to corner the world market on demand data.
Don’t just sense demand, understand it
In Amazon’s early days, mega-retailers like Walmart, Tesco and Best Buy were seen as all-powerful because of their grip on consumer demand. For years, e-commerce was dismissed as a specialty side show, suitable for books and maybe electronics, but little more. That illusion is long gone.
The winning formula was, perhaps ironically, little different from what retail has always offered: convenience, selection and price. Amazon, of course, built a spectacular supply chain to bring this value proposition to shoppers’ homes, but the real secret sauce lay in its mind-boggling ability to link item-level demand to lifestyle-level need. Consumers know they’re not anonymous when shopping on Amazon, so suggestions on what else to buy, when to join Prime for “free delivery”, or whether to allow an inanimate, but intelligent agent (Echo) into your home all seem perfectly reasonable.
Amazon is not just sensing and responding to item demand so much as painstakingly developing a complete picture of each customer’s demand curves. The personal data being collected by Amazon can certainly be used to know the difference between ‘must-haves’ and ‘nice-to-haves’. For supply chain strategists, the implications are clear: improve your ability to sense demand, but don’t forget to also develop your understanding of what it really means.
Our benign overlord
When Bezos specified “training data” as a key to the development of artificial intelligence, he meant real-world case data of our shopping preferences. If applied narrowly to a single product or even category, AI could certainly help sharpen pricing and promotions.
What Amazon is positioning itself to do, however, is far more ambitious and involves what AI experts call ‘contextual awareness’. This means knowing not only the what, but also the when, why, where and how of consumer need. The long game is all about selling us not just what we want, but what we need, and probably before we realise we need it.
I’ve often been suspicious of Amazon for its demonstrated hunger to know everything about us. Increasingly though, I believe Bezos when he says the company thinks of itself as tied to customers, not specific technologies.
Combining technologies for sense and respond with others able to understand could bring breakthroughs not only for consumers but for society.