With judicious investments, Dell aims to stake a sizable claim in AI territory
Going into the Dell Annual Analyst Summit, held this year in Chicago, and featuring the theme of artificial intelligence (AI), I had only two questions: one, why does the company keep messing with its cap table? And two, why Chicago?
Right in the first session, Irene Mirageas, worldwide head of analyst relations, spelled out in no uncertain terms that there would be no discussion of ownership changes, stock floatations, and the rest the liabilities side of the balance sheet. Anything on technologies, products, go-to-market, channels, marketing themes, branding, and general market developments, fine. Just don’t ask about the man behind the curtain.
That settled, why Chicago? took on greater urgency as an icy wind blew into the city off Lake Michigan, emphasizing the need to better understand this choice. The old cheap-and-convenient-hotel-and-conference-space riff didn’t really cut it. I mean, 300 analysts is a lot of analysts, but, in the scheme of things, not a lot of actual people. Not like the Consumer Electronics Show (CES) crowd that clocks in at nearly two hundred thousand. Logistics were not really a hurdle here. One senior manager speculated that direct flights from Asia were convenient to O’Hare, which seemed plausible, as we had a small but significant cadre from over the pole. Nonetheless, I remained unconvinced. Finally, someone uttered the cryptic phrase “executive convenience,” which did sort of explain things.
The top brass all showed up. In the comfortable midday hours, founder and CEO Michael Dell held forth in his usual manner, taking questions easily. Executive Chairman Jeff Clarke gave a masterful performance, strutting and fretting his hour upon the stage. But I found it difficult to imagine what makes Chicago more or less convenient for people who have their own fully tricked-out transport.
I began to form a theory of convenience something like this: there are perhaps a dozen-to-a-score of events that the company holds each year in each region of the world and sprinkled around the various sections of the United States. Perhaps, turning up at each location honors an important set of customers. It’s entirely possible that the analyst summit was a prop within a larger play. Like Groucho before me, I’m suspicious of any club that would have me for a member.
Or perhaps the leadership was in town to discuss the cap table with key players in the financial community, although that would tend to argue for New York. Likely, we’ll never know.
As to the formal program, Scott Darling, president of Dell Technologies Capital, led off with a rigorous exposition of the company’s investments in startups, elucidating a philosophy that focuses on areas where the company has deep domain expertise and a strategic interest in the outcome. In a typical venture, Dell takes on high technical risk, leads investment rounds, takes board seats, and is generally activist. Darling illustrated the breadth of the company’s investments in AI by bringing on a roster of speakers from five portfolio companies, which represented a nice ladder up the stack from silicon, to data, to network security, operations security, and finally applications.
Nigel Toon, CEO of Graphcore, talked about a 23.6 billion transistor microprocessor the company is making that has 300MB of on-chip memory, allowing for much-accelerated AI operations. Matt Carroll, CEO of Immuta, talked up his firm’s “dynamic differential privacy” policy platform that takes into account the interests of the data owner, data scientists, and lawyers when using masses of data for AI purposes. May Wong, CTO of Zingbox, described how her company helps secure large network-attached medical equipment — essentially complex Internet of Things (IoT) nodes — by monitoring its network traffic behavioral for anomalies. Greg Martin, CEO of JASK, explained how his company’s software uses AI to filter the millions of alerts that bombard a typical big-company security operations center, highlighting visually the ones most likely to be dangerous. Stephen Pratt, CEO of Noodle.ai, gave a solid example of how his company’s AI application helps make huge arc furnaces more efficient at Big River Steel.
After top management laid out a remarkably uniform picture of how it is funding a whole series of efforts to integrate AI into its products and services, Karen Quintos, Dell’s Chief Customer Officer, sat down with three customers who are implementing AI in their own environments.
That’s when it got really interesting. All of them do major business with Dell, largely hardware, but also software and services. Carnival Corp.’s Greg Sullivan, VP of information technology, explained how the Miami-based cruise company uses AI. And AI is looming larger on the screen of Jack King, chief technology officer in the Department of Innovation and Technology for the State of Illinois, which looks after things like transport, infrastructure, the lottery, school policy, libraries, employment, and myriad other services that the state provides its residents. Mastercard, based in Purchase, NY, is known as a major credit card company, although its VP of global big data consulting, Nick Curcuru, took pains to point out that the plastic card is an anachronism. These days, customers can do financial transactions with a key fob, phone, wearable device, or even a car.
I was particularly taken with the picture Carnival’s Sullivan painted. Although I have never taken a cruise myself and have rather a horror of the idea of being cooped up on floating hotel with a bunch of other revelers, his examples really drew out the ethical issues related to AI. A lot of data, some of it quite sensitive, is associated with each customer. This data can be used to enhance the customer experience, for example, by offering choices based on past decisions. It can also be used in aggregate to figure out when to have the maximum number of food stations open during the dinner buffet.
Sullivan explained how data is managed before, during, and after a cruise. Essentially, a new record is set up for each trip, the data is loaded onto the ship’s computers (sometimes via “sneakernet” sometimes over WiFi, sometimes via a long cable, depending on that particular port’s infrastructure), and then it is pulled back off after debarkation. This data includes the passenger manifest as well as the ship and crew log.
Passengers can opt into the OceanMedallion program, through which they are issued a coin-size token that can be put in a pocket, worn as a pendant, or attached with a clip. The device, which has no on-off switch or any discernible technology, uses near-field communication and Bluetooth to interact with shipboard features and functions. People can buy goods and entertainment with it, pay for food and drink, and even open their cabin doors, all without having to touch anything or deal with any paper. This convenience is great, but the device also can be used in conjunction with on-board cameras to track people as they move around the vessel. While location services are fine for helping people navigate, they can also show that the gentleman from 304 on the third deck dropped down to the cabin of the lady in 206 on the second deck at 2 a.m. for a tête-à-tête.
So, even this early on in AI implementation, Carnival has come right up against the ethical issues of gathering huge amounts of cross-referenceable data. And more data is coming. Next year, the company will start deploying cabin-door cams. Sullivan explained that only the data for a particular sailing goes onto a ship. Ships are lit up with 800Mbps of WiFi, which passengers immediately fill with traffic. When the cruise is done, the data goes ashore, where it is retained for legal reasons for a period of time. Then, it is anonymized, aggregated, and fed to AI programs with the aim of improving shipboard services by better understanding passenger behavior.
In a nod to the thorny nexus of big data, AI, and privacy, Curcuru told the audience that Mastercard has an ethics committee composed of senior executives to make sure policies are in line with company goals. King also mentioned that Illinois has focused “a lot of scrutiny on privacy issues.”
AI will enable many new and amazing things if data privacy is handled properly. Otherwise, it could be an Orwellian nightmare.
Whichever way it goes, Dell will be there at the forefront, developing its own AI and nurturing that of its customers. Big data isn’t going away. It’s only getting bigger, and Dell intends to be at the heart of the business of letting all this information tell us its story.