By Aara Ramesh
In January 2017, one 20-day tournament sent ripples of shock through the tech industry, when Libratus, an AI-powered robot, beat four of the world’s best poker players in a series of no-limit Texas Hold ‘Em games. This was seen as particularly significant due to the fact that poker is a game with imperfect information, where the AI has to be able to understand and interpret bluffs from competitors and be able to bluff itself. This has long been considered something that AI would never be able to do — the factor that particularly elevates human intelligence over computing systems’ intelligence.
Libratus’ shock win set off a longer-term discussion about the role that the latest technology can play in the gaming and gambling industries and what this could possibly mean for the sector’s future.
Artificial Intelligence (AI) is the use of technology to enable computers to replicate human behavior in terms of visual perception, speech recognition, and decision making. The AI algorithm looks for and identifies specific patterns and links over thousands (if not trillions) of interactions with humans, and then can choose its own strategy or response based on that analysis. It does not actually possess and exercise the knowledge bank in the traditional sense; rather, the algorithm just runs through data and analyzes it.
Gaming itself is somewhat of a treasure trove of the type of data that is suitable to train AI algorithms, as player behavior, choices, and game strategies can all be examined. While humans can also trawl through and analyze data, that process could take hours. Computer systems are able to much more quickly sift through millions of data points, enabling greater efficiency.
With AI being such a nascent field in general, the gaming industry is still exploring and understanding how it can best be utilized. Let us take a look at the field as it stands and at some of the potential uses of AI in casinos.
The State Of The Online Gaming Industry
Though exact estimates vary based on who commissioned and conducted the study, the overall casino and gambling industry — both land-based and online — is thought to have been worth roughly $227 billion in 2020. The market size of global online gambling alone is believed to have been anywhere between $57 to $59 billion last year. This is expected to double within the next five or six years, to be worth between $92 and $97 billion by 2026.
In the U.S., the online gaming market was valued at around $1.98 billion in 2020 and is expected to grow by more than 17% between 2021 and 2026. With more and more states in the U.S. legalizing sports betting after the landmark Supreme Court decision to reverse the ban on it in 2018, the market is set to grow in leaps and bounds in the short term.
Security, Fraud, And Cheating
With this growth and the introduction of new technologies comes fears over security, fraud, and cheating. It seems like that for as long as there’s been gambling, there’s been people who try to exploit the system for their own gain. That didn’t change much with the incorporation of technology and the internet into gaming.
For its part, AI both attracts and thwarts cheaters. In land-based casinos, it can often be difficult to monitor every single inch of the floor on an ongoing basis. The use of AI can identify (based on parameters established by the operator) instances of theft and cheating happening in real time, and can alert security personnel, who can then take appropriate actions.
Security teams are not able to access cameras or surveillance equipment to monitor suspicious behavior from a person gaming in their own home. AI, however, can help in the large-scale analysis of patterns and actions that could identify cheaters and help betting maintain the integrity of the industry.
AI can also help in a regulatory capacity, by automating and optimizing the systems that monitor a player’s age and location, which would determine the legality of them engaging in gambling.
However, there is a risk that the use of AI is not entirely a deterrent. Cheaters can develop their own AI technology to battle the casinos. At the very least, however, this helps even out the playing field and the problem can be mitigated by keeping the technology up to date and investing in equipping it with all the information it needs to continuously refine its algorithm.
Heading Off Potential Gambling Addiction
With the growth of the online gaming industry comes renewed fears about potentially unhealthy gambling behaviors taking root. But this too can be — and is being — tackled through AI employed by casinos and online betting sites.
AI presents great potential to help curb destructive gaming behavior. The vision for the use of AI in this respect is that technologies can analyze a player’s actions, like how long he or she is playing, how much they’re spending, when they’re spending, what games are they spending on, how much they are staking, etc. Once the algorithm recognizes and maps patterns in specific users, it can identify when someone is veering towards irresponsible gambling.
After this, the technology can then lock down the system before any serious loss is incurred, and can inform both the players and the game operator. The latter can also set loss limits and automatic timeouts for certain types of behavior, and suspend accounts of those displaying risky habits.
One such solution is BetBuddy, which was acquired by Playtech in 2017. The platform was created through the joint efforts of AI and machine learning experts, and clinical researchers looking into unhealthy gambling patterns. The result is a piece of predictive technology that is thought to be able to detect a potential gambling addiction with an 87% rate of accuracy. BetBuddy’s predictive capabilities make it an industry-leader in implementing responsible gambling practices, according to its creators. They say it is continuously updated based on the latest research so that it remains cutting edge.
Similarly, in the UK, the betting industry proactively implemented in 2019 the Anonymous Player Awareness System at fixed-odds betting terminals. This technology was trained to analyze player behavior and spot troublesome patterns like “chasing losses, playing erratically, speeding up betting while losing, or simply spending too much time on the machine.” If it did detect such behavior, it would automatically shut down to impose a “cooling off” period of a minimum of 30 seconds, for the player to step away and calm down.
AI tools in this setting can better equip players, operators, and regulators with the information necessary to head off the problem. Spotting early warning signs can be crucial to preventing addictions before it becomes a crippling or overwhelming problem.
Growing The Market
Today, AI systems are being used in a variety of settings to enhance the user experience. This is applicable to the gambling market as well. Just as the introduction of internet-based casinos and gambling has grown the overall market, so too can AI help in doing the same.
Not only can AI manage and set specific odds and pay-outs, it can allow casinos to give players a more customized, personalized experience. This includes better deals and special, tailored offers that can serve as a huge draw for potential and returning customers, likely increasing the revenue of the market.
Better, faster, and smarter use of data can help companies understand and anticipate a players’ needs, by figuring out what entices them to play, why they quit, and why they may change games. All these mechanisms that are designed for a better game for players could help drive the industry into the next phase of its expansion.
What Happens Next?
The fact is that AI is the future for humankind — not a replacement, but an aid, a support system to better our lives.
Let’s go back to the robot Libratus, who trounced its human opponents and won a hypothetical $1.5 million in 2017.
Much like with any other form of technology, the algorithm behind Libratus is not restricted to being applied in gambling or poker alone. Having demonstrated an ability to strategize and negotiate, the software can be implemented in a variety of fields, including in private-sector transactions, the military, cybersecurity, and healthcare. Basically, any situation that calls for strategic reasoning amidst conditions of imperfect information is ideal for this technology.
So is its destiny to be a poker maestro forever?
No. Today, Libratus is an employee of the Department of Defense, where it is engaged in an unspecified role in the Defense Innovation Unit. And while its opponents were left exhausted and reeling over Libratus’ triumph, it was not all bad news for those top-dogs of poker. “It’s almost like we’ve been shell-shocked into being much stronger players,” one of them said. “Nothing anyone does will seem that crazy anymore.”