Two Questions for Managers of Learning Machines

Two questions that managers of intelligent machines should ask: It’s been a couple of years since Stephen Hawking warned that artificial intelligence could “spell the end of the human race.” The terminators aren’t here yet and unless they come very soon, the managers of AI-based technology have a couple of more immediate issues to address, according to Vasant Dhar of NYU’s Stern School of Business and Center for Data Science.

The first, which Dhar takes up in a new article on TechCrunch, is how to “design intelligent learning machines that minimize undesirable behavior.” Pointing to two high-profile juvenile delinquents, Microsoft’s Tay and Google’s Lexus, he reminds us that it’s very hard to control AI machines in complex settings. “There is no clear answer to this vexing issue,” says Dhar. But he does offer some guidance: Analyze the machine’s training errors; use an “adversary” — through means such as crowdsourcing — to try to trip up the machine; and estimate the cost of error scenarios to better manage risks.

The second question, which Dhar explores in an article for HBR.org, is when and when not to allow AI machines to make decisions. “We don’t have any framework for evaluating which decisions we should be comfortable delegating to algorithms and which ones humans should retain,” he writes. “That’s surprising, given the high stakes involved.” Dhar suggests addressing this issue with a risk-oriented framework that he calls a Decision Automation Map. The map plots decisions in two independent dimensions — predictability and cost per error — and suggest whether it would be better made by human or machine.

The DAO of blockchain business: In the middle of its month-long crowdfunding campaign, a new start-up company named DAO had raised the equivalent of $150 million in ethers (ETH) digital tokens. As yet, DAO has no products, no services, and thus, no sales — except for bits (bytes?) of itself. So, what are investors getting for their money? An ownership stake in a business structure built on blockchain technology.

A $150 million stake seems like a lot of money for what Michael del Castillo describes in CoinDesk as a “distributed organization with no single leader that could theoretically exist so long as there’s an Internet connection [that] was launched last month, and has since then left many observers and Ethereum community members feeling optimistic — if not a bit confused — about what exactly was created.” But who knows? Maybe it will be the deal of this century.

What’s interesting about DAO is its blockchain-based business structure. There’s no leadership team, except for a group of so-called curators. Investors in DAO are buying digital voting rights, which they will use to direct funds to proposals submitted by entrepreneurs. If they vote well, the successful businesses they’ve funded will make them rich. That’s the pitch: Ownership in a company that aspires to be Berkshire Hathaway sans Warren Buffett and Charlie Munger.

Unsurprisingly, many observers think DAO is the emperor’s new clothes. Bloomberg View columnist Matt Levine suggests DAO is “a centuries-old idea [that] has been dressed up in cryptographical mystification.” And Tom Simonite, the San Francisco bureau chief for MIT Technology Review, thinks that “like so many ideas buzzing around the cryptocurrency world, DAO probably won’t live up to its own hype.”

Whether you see DAO as a company without executives or a VC fund without VCs, it’s an experiment in digital organization structure that’s gonna be interesting to watch.

How not to PARC an innovation lab: In the 1970s and early 1980s, Xerox’s Palo Alto Research Center (PARC) developed many of the mainstays of personal computing. Unfortunately, there was a disconnect between PARC’s breakthroughs and Xerox’s core copier business. So, PARC ended up helping companies like Apple and Microsoft more than its own parent.

Today, innovation labs are back in style. But they’re not cheap, and no exec wants to repeat Xerox’s mistakes. To avoid that fate, the corporate overlords of innovation should take a few minutes to read a new article for ZDNet by Mark Samuels.

Samuels reports that innovation labs should be operated independently of business units, but aligned strategically. “Labs for the sake of innovation are completely wrong,” Mark Ridley, director of technology at job site Reed.co.uk, tells the business journalist. “Firms should not create a separate innovation strategy. There should just be a business strategy, which includes your aims and objectives for innovation.”

Reed.co.uk’s Monday Labs lifts small teams out of the core business and gives them a few months of uninterrupted time to work on an idea for a new business. Each team, explains Samuels, is composed of “five key archetype personas: hustler, hacker, creative, designer, and visionary. These cross-startup roles covered key activities, such as management, sales, and development.”

“We’d forgotten just how powerful the right four or five people in a room can be,” Ridley tells Samuels. The result: Monday Labs has spawned three spinoff businesses, one of which is now an established arm of the main business.