In the past few years, we’ve witnessed artificial intelligence (AI) systems beat human champions in Go and Jeopardy! But earlier this year, AI tackled a new frontier: Poker, specifically no-limit Texas Hold’em.

The AI system, Libratus, became the first AI to beat professional poker players at heads-up, no-limit Texas Hold’em during the “Brains vs. Artificial Intelligence” tournament. But the benefit of AI systems that can play poker is a breakthrough that’s about much more than just a game.

“Poker requires reasoning and intelligence that has proven difficult for machines to imitate,” writes Will Knight in the MIT Technology Review. “It is fundamentally different from checkers, chess, or Go, because an opponent’s hand remains hidden from view during play,” a concept known as imperfect information. “In games of imperfect information, it is enormously complicated to figure out the ideal strategy given every possible approach your opponent may be taking,” Knight writes. “And no-limit Texas Hold’em is especially challenging because an opponent could essentially bet any amount.”

Libratus’ capacity to learn the rules of poker and adapt to unpredictable human strategies highlights its versatility: “The Libratus algorithms can take the ‘rules’ of any imperfect-information game or scenario and then come up with its own strategy,” writes Jeremy Hsu in IEEE Spectrum. That has business implications that alter the rules of engagement across industries:

Bidding: Learning the rules of the “game,” and then learning rational ways to exploit them, also has business applications such as bargaining and auctions, writes Chris Velazco in Engadget. “When the FCC auctions spectrum licenses, they sell tens of billions of dollars of spectrum per auction, yet nobody knows even one rational way of bidding,” Thomas Sandholm, professor in the computer science department at Carnegie Mellon University, tells Velazco. “Wouldn’t it be nice if you had some AI support?”

For example, Google is using artificial intelligence in bidding for online advertising, writes Katherine Noyes in PC World.  “Tapping some of the same artificial-intelligence technologies that have already appeared in Google Photos and AlphaGo, Smart Bidding is a new capability for conversion-based automated bidding across AdWords and DoubleClick Search to help companies determine their optimal bid for any given campaign or portfolio,” she writes. “It can factor in millions of signals, Google says, and continually refines models of users’ conversion performance at different bid levels.” The software does this by evaluating patterns in areas such as campaign structure, landing pages, ad text, product information, and keyword phrases, the company explains.

Bluffing: The fact that AI can now strategize is “what could make it useful in a board room, or a war room,” writes Dave Gershgorn in Quartz. “The software looks at its situation, and tries to predict the best move for nearly every potential way the game could play out, based on games it’s played before. When the opponent makes a move, the calculations are made again, with the opposition’s decision informing a narrower set of potential outcomes. In poker, that means maximizing the chance of winning while not knowing the opponent’s cards. In the real world, it could mean knowing when a foreign power is bluffing about military strength.”

Healing: In healthcare, AI could use game methodology to look at an infection as “the opponent” and taking sequential actions and measuring the results, could be the “game,” Sandholm explains. Such algorithms could also be used to help manage chronic diseases such as cancer by optimizing the use of existing treatments as well as developing new ones, he continues.

Breakthroughs Boost Investment in AI & Raise New Questions for Talent Management

The skills learned by AI systems such as poker-playing is what’s driving a lot of investment in AI by enterprises, predicts Forrester in its report Predictions 2017: Artificial Intelligence Will Drive The Insights Revolution.  After noting a greater than 300 percent  increase in investment in cognitive computing in 2017 compared with 2016, the report concludes that “AI will drive faster business decisions in marketing, eCommerce, product management, and other areas of the business by helping close the gap from insights to action.”

At the moment, technology that can beat the pros is limited in its user base, but that won’t last. Libratus is housed on Bridges, a supercomputer at the Pittsburgh Supercomputing Center. That doesn’t exactly sound like a technology poised for mass adoption anytime soon—but more than 2,500 people have already used Bridges for 650 projects. Nick Nystrom, a senior director of research at the Center, stresses the accessibility of its technology to non-experts. “It’s a very cloud-like model letting people who are not programmers, not computer scientists, not supercomputer users make use of a supercomputer without necessarily even knowing it.” So already, people without a technical background can hire out human-beating talent to achieve specific ends.

That means AI has already crossed into the territory of talent management strategy. That’s especially clear in scenarios where humans and AI work side by side. Consider the role AI may play in dealmaking when sensors could determine physical aspects in an opponent such as “pupil dilation, mannerisms, the amount a player is sweating and other biological signs of stress (bluffing) to empower its decision-making,” writes Kay Firth-Butterfield, an attorney and executive director of a nonprofit organization working on the uses of AI for good, in the Guardian. “If we transfer this skill set into business, military, government and diplomacy, AI, possibly embedded in a robot, becomes an invaluable aid in negotiations to assess whether the negotiator on the other side has a strong or weak position.”  Sales teams have long been able to point to closed details as the proof of their tactics to persuade and sway—but what happens when AI can furnish a counter-narrative about what works, and what moments were really decisive? Where those conflict with human judgment, what should carry the day? Similar questions seem to wait in the wings for corporate legal and compliance professionals as this technology is tapped to solve problems in health, safety, and risk management.

Surprising though it may seem, this points back to the need for human capital. For all the talk of AI being free of human bias, it must be acknowledged that AI can take on a bias all its own, and that this often is acquired through the data sources and interactions ‘fed’ to them (Think of how quickly Microsoft’s ‘Tay’ Twitter chatbot was taught to be racist and misogynistic). Observing just these points, Kristan Hammond, writing in TechCrunch, argues for the creation of a human ‘mentor’ for AI to monitor for emerging biases. Elsewhere, Scott Middleton writes how his company, Stratejos, is already developing AI that can link to lines of code and the person who wrote them, in effect creating a project management tool. This raises the question of the ultimate end goals against which performance is to be measured.

And this takes us to the oft-cited benefit of AI, its potential to free up executives for more creative and strategic thinking. Even assuming all goes according to such rosy predictions, the question of how best to use this time remains. Google’s famous ’20 percent time’ reserved for engineers to innovate was eventually phased out around 2013 as the company’s growth and earnings expectations created pressure for a more focused approach to innovation.  That’s all to say that ‘getting time back’ using AI sounds great, but as the Google example shows, the value of time for innovation is inseparable from the broader strategy it’s meant to serve. So there needs to be structured plans in place on how time freed by AI will be used that map to well-defined objectives for the company  in order for it to be a success.

Far from eliminating humans, these developments elevate the importance of human judgment and the need to build real-world understanding of AI across divisions.

What it all means: A 3-part strategy for executives

What was once a futuristic notion is fast becoming reality. A 2016 survey showed that 38 percent of organizations already employ artificial intelligence, and that by next year, that percentage is expected to increase to more than 60 percent, writes Gurbaksh Chahal in HuffPost. CIOs can help their companies prepare for and leverage the technology with a 3-part strategy:

  1. Start at the top: Get buy-in from the C suite and emphasize the need to re-evaluate AI strategy and objectives. The exact articulation is key, as it will determine the basis for how highly literal bots will run and how they will interface with human mentors and teams alike.
  2. Re-think roles and build and an AI-aware culture: Think of teams’ composition, with an eye for ensuring a distribution of technical expertise in managerial and HR roles. People across different lines of business must be prepared to interface, and make judgments on the actions of AI. In recent years, Apple has gone on an AI expert hiring spree. Yes, AI and machine learning experts can be difficult to find and expensive, but hiring experts in computer science can be a first step towards developing a culture of comfort acting on real-world AI concepts rather than fearing science-fiction fantasies.
  3. Set expectations for a different kind of company: CIOs must work to set expectations with stakeholders to allow them to pursue AI-based opportunities, such as greater time for innovation.

Executives are fortunate that technologies such as Libratus still not widespread, meaning that there is time to act. But perhaps only just. Now is not the time to squander being dealt a good hand.


Simplicity 2.0 is where we examine the intricate and transitory world of technology—through a Laserfiche lens. By keeping an eye on larger trends, we aim to make software that’s relevant to modern day workers, rather than build technology for technology’s sake.

Subscribe to Simplicity 2.0 and follow us on Twitter. If what we’re saying piques your interest, head over to Laserfiche.com where you’ll see how we apply the lessons learned on Simplicity 2.0 to our own processes, products and industry.

Automation

Robotic Process Automation, or RPA, works by teaching a software robot how to work with existing software to perform a process. Listen to learn more.

Listen Now

Related Articles

By Sharon Fisher, May 18, 2016

As far back as Gutenberg, leading-edge technology has been used with the Bible. Now scientists are using AI for Bible interpretation.

Read More

By Sharon Fisher, November 09, 2016

Drowning in irrelevant search results? Artificial intelligence could help produce "intelligent search" by making search smarter.

Read More