Artificial intelligence or AI is a hot field of research right now. And there is a lot of hype and buzz being created in the public realm as well, given how fascinated we all are about killer robots and murderous AI. We may have Hollywood to thank for such a negative perception about AI in general.
But killer robots aside, the implications from the rise of AI are all too real. While it could be very helpful in many industries, AI could easily put a lot of people out of business. If the fears of some experts are real, that could include online casinos as well.
Understanding Casino RNG
Both AI and online casinos occupy a common ground – both exist in the digital realm and owe their entire existence to software code. All modern casinos use software for their most popular games – slots. Even the machines inside bricks and mortar casinos use software to run the games.
And the entire i-gaming industry relies on the continued success of this particular software, which is called a Random Number Generator, or RNG. A casino makes money by hosting games of chance, where neither the player nor the host casino can accurately predict the outcome.
Even in these random games, the casino holds an advantage known as the ‘house edge’. It is linked to the famous saying about casinos – “no matter what, ultimately, the house always wins.” This is actually a fair mechanism – after all, casinos are businesses too and they need to turn a profit. The house edge is like the fee a player pays to the casino.
And in online casinos, the RNG software is responsible for ensuring that games remain random, or unpredictable. The software uses complex algorithms that are constantly calculating outcomes, even when the machine is not being played. In this way, the results of consecutive plays are never linked.
But if the casinos design the software, can the results be trusted? This is where regulatory bodies like the UK Gambling Commission and industry bodies enter the picture. They mandate that the RNG software behind all casino games be tested at independent labs to ensure that they remain fair and not rigged excessively in favor of the casino.
Can RNG Software be Predicted
There is a reason why casinos protect their RNG software like the Holy Grail. If the software falls into the wrong hands, it could be “cracked,” leading to serious repercussions for any casino running that software.
It has already happened in the industry, as highlighted by this story that ran in Wired a few years back. A long intriguing read, it tells the story of Alex, a Russian math and programming genius who reverse-engineered the RNG software inside major western casinos.
The information he gained by decoding the algorithms enabled Alex to correctly predict when a slot machine is going to payout. He used this knowledge to make millions of dollars targeting slot machines all over the world.
Instances like the one involving Alex can be prevented to an extent. Casinos just have to take extra care to keep the algorithms safe. And in the event one does get “lost,” they could just decommission it and get another one. Those are expensive measures, but still better than getting bled dry by the hackers.
How AI Could be More Dangerous than Humans
Though they call it “Random Number Generator,” the correct technical term for the RNG software found in modern casinos is Pseudo RNG or PRNG. For the most part, creating a truly random number generator is incredibly complex and expensive. It can be done, but you would need to base it on natural phenomena, like the decay of a radioisotope.
You cannot use that kind of mechanism in casino games and run a profit. Instead, casinos develop long and complex mathematical algorithms to create RNG that appears random to humans. These RNG do exhibit patterns that can be measured and predicted, but doing that would take humans a long time indeed.
But AI is a different kind of intelligence. Modern advances in creating AI are centered on things like big data, machine learning, and neural networks. All these have ominous implications for casinos and RNG.
For starters, the output of an RNG system over a period of time like weeks or months can include millions of calculations and results. It is a textbook definition of big data, which are basically massive volumes of data that would overwhelm humans.
But AI thrives on big data. Machine learning is a technique used to teach AI to do things. It involves feeding tons of data and specific instructions. The AI combines the two and seeks out patterns hidden in the data. Once the AI has built up a huge enough collection of patterns, it can then reliably make predicts related to that data.
Does that sound familiar to anything we touched upon earlier? If someone creates an AI and instructs it to study the outcomes spewing from casino software, it would be just a matter of time before the AI manages to learn the pattern. Since machines don’t get tired, the AI could crack the RNG in a very short time indeed.
How will the casino industry respond to this threat?
There are no easy answers to this question. One option might be to get smarter, more complex RNG software, possibly those based on AI themselves. There is no reason why it has to be all doom and gloom for the casinos.
The rise of the AI promises a lot of benefits for the casinos as well. There is already much talk of using AI to run casinos on their own, dealing with customers using chatbots, and using predictive analysis to give players a more personalized experience.
After all, human ingenuity is not just about developing new technology, but also about harnessing it for both offense and defense. If the other side uses tools to cheat, casinos too can play the same game. The cat and mouse game between casinos and cheaters is as long as the industry itself. The AI may add a new edge to it, but it will probably not end the game for good.