Author(s): Martin Garner
Amazon's inaugural re:MARS public event in Las Vegas has just finished. It developed from a private conference that Amazon has run for several years, and continues to do so. The agenda covered machine learning, automation, robotics and space (MARS) and had a very broad scope. It ranged from technical sessions about machine learning — for example, giving details about the technology behind Amazon's cashierless Go stores — to ways of saving the planet by using technology better.
Most technology conferences are fairly focussed, so Amazon took considerable risk with such a mixed agenda. But the risk paid off. Who wouldn't be interested in a session titled "Fighting human trafficking using machine learning" or "Solving Earth's biggest problems with a cloud in space"? Although the event was emphatically centred on technology and primarily aimed at developers, it attracted a diverse audience including professors, astronauts, actors and CEOs. It had about it a slight feel of the World Economic Forum's meetings in Davos.
It was also a rare glimpse into how things work in Amazon. Two things stood out in this respect: the company's scale and culture. The scale is astonishing — Amazon is happy to take on projects like creating its own version of Google Maps, with details down to 5 mm, to enable deliveries in cities using small six-wheel robots through Amazon Scout. The company also forecasts demand for 400 million products, with variations in regions and delivery methods, and updates these forecasts daily.
It's useful to know that Amazon's revenue over the past four quarters is equivalent to the annual GDP of the Czech Republic, the 44th largest country in the world by GDP, with a population of nearly 11 million people compared with Amazon's 630,000 employees. More than that, the Czech Republic is one of the fastest-growing economies in Europe, with GDP rising at about 4.5% per year; Amazon grew on average 25% per year during the past four quarters.
With that kind of scale and growth you'd expect a degree of swagger or even arrogance about the senior staff, but instead the culture comes across as an interesting mixture of humility and ambition. Clearly people in Amazon are well aware of the extraordinary achievement in building the company, but they also seem to be grounded by a strong sense that many of the areas they're working on are so new that there's no playbook. Everyone's finding their way and, of course, not all developments will go smoothly. It was refreshing to find that a large public company could be so candid about that, especially at an event with many members of the press on hand.
But Amazon's extremely ambitious facet also comes through, in various ways. Partly it's that whatever it develops next has to be big or it won't move the needle on the business. Partly it's that the enormous computing power available with cloud services, together with artificial intelligence, enabling organizations to tackle complicated issues that they've been unable to address before. And partly it comes from a willingness by Amazon to think differently. For example, CEO Jeff Bezos mentioned his view that heavy industry should operate in space, or on the moon, as soon as possible to leave Earth less crowded and polluted. This is one of the main tenets of his venture into reusable rockets through Blue Origin.
There was a strong sense from Amazon and all other speakers that artificial intelligence is foundational to the next phase of human development. It brings an opportunity to build our expertise, just as the Internet expanded our ability to move information around the globe. A leading purpose of the conference was to position Amazon as a thought-leader in the area, a company with a long history of using artificial intelligence at large scale and one that's fully plugged in to academia, ventures and individuals who are using the technology to try to improve the world.
Few would disagree with the importance of artificial intelligence, but one of the risks with this technology is in appearing to put most of the effort into the technology and perhaps not enough into addressing societal issues raised. It was clear that Amazon is acutely aware of those issues, and is working on them, but is perhaps less confident in talking about them, preferring external speakers to take that role. Contrast this with Microsoft, which is taking a strong stand on some of the major social aspects of artificial intelligence.
Amazon's general position is as a company that does the heavy lifting needed to provide the infrastructure and services that make it possible and manageable for others to take on huge data and challenges in artificial intelligence. This means it may be behind the scenes in those projects, and not necessarily visible as an active agent.
However, Amazon's scale means that it will be both a poster child and a whipping boy when things go wrong. So, we suspect that the solid commercial principles, including "customer obsession" and "passion for invention", that have helped build the company to what it is today may not be enough for the future. Perhaps those values will need to rise to a higher level to reflect Amazon's increasing involvement in shaping society.