Although it’s often seen as a stable part of futuristic Sci-Fi films, Artificial Intelligence is nothing new.
Entering the gambling industry in blistering style in the 1960s, an AI programme developed by cybernetic scientist Arthur Samuel managed to beat the best American checker player at the time. Since that victorious entrance, the technology has evolved to compete against human players in various different games, with programmers constantly aiming to create software capable of beating the world’s best.
As with Samuel’s pioneering robotics, they’ve often succeeded, although some have taken longer than others. Take poker, for example, one of the industry’s most popular games and potentially its best-known. Although many have tried their hand at creating bots for this notoriously difficult game – where skill and chance are interwoven in an incredibly complex web of probabilities and random outcomes – it took until as recently as 2015 to create the fully-fledged supercomputer ‘Claudico’.
Yet Claudico wasn’t perfect and, to the distress of many hoping to make millions online with minimal effort, neither are any other bots. In fact, most are downright amateurs still in the very early stages of development and created mainly for personal interest by people who don’t possess the mathematical genius to turn them into formidable threats. Some may be able to beat beginners, but the vast majority would fall down if faced by a professional.
Even then, their chances of victory are far from guaranteed. This is because the element of chance throws up numerous unknown variables that are extremely difficult to account for. Compare poker to chess, for instance: whereas a computer can know all possible moves its opponent could make on the board, it can’t know for sure which card will come out of a deck next. Not to mention you can see your opponent’s pieces on a chessboard, whereas you can’t know what your fellow poker players are holding, or to what extent any are bluffing.
So what’s all the fuss about? In a nutshell, advancement. Technology has been improving at an astonishing rate in all areas over the past couple of decades and will almost certainly pick up the pace over the coming ones. Players, both professionals and amateurs, are right to be wary of dismissing the current infancy of many AI bots, as even if they’re capable of being bested right now, they’re set to get better over time.
This will lead us to the point where, ultimately, even the most gifted human cannot maintain the same number of card combinations and possible manoeuvres in their head as a computer can, nor can they do mathematical calculations at the same speed and depth as this adversary. In short, we just aren’t as clever as computers and technology will inevitably reach a point where, as long as there’s an element of strategy to the game, the odds of victory are going to be strongly in their favour.
And that’s just focusing on the professionals, who make up a tiny proportion of the world’s online gamers. The vast majority are amateurs and part-timers, playing for fun or practice. Even many of the under-developed AI bots in play today are capable of raking in the winnings against these lesser opponents, a fact that could drive them away from the online scene. Without their revenue, many sites would struggle to keep going.
It would be comforting to assume that security measures could keep such bots at bay, and in many cases, they can. But if a bot manages to slip past software detectors, moderators then have to rely on statistical data and non-conforming habits to spot any intruders. Success at this point would be largely dependent on luck, as programmers could design bots to occasionally lose and only aim for small yields, avoiding the red lights that would start flashing if a new player immediately won thousands. As well as that, professionals tend to act very organised and not interact with others while playing, meaning that distinguishing any non-human behaviour is difficult when even the humans are exhibiting it.
Finally, even if AI bots never become good enough to do all of this – or simply get caught too often by security systems to be a significant threat – widespread paranoia over them could drive players away regardless. After all, why take the risk?
An Overblown Fear
Yet all of these fears are ultimately unlikely to materialise, at least not to the extent of industrial armageddon. For years, players have been returning in their limitless numbers to casino games that are already rigged in the House’s favour, so it’s unlikely that the minute possibility of encountering an AI bot would deter them from gambling online now, especially those who were mainly playing for fun anyway.
As well as that, the severity of the threat AI bots pose largely comes down to who has the best technology: the invaders or the site. Wheres smaller companies will be vulnerable to well-funded, well-organised crime rings, the latter will struggle to compete against the endless resources and manpower large corporations can throw at their security operations. A few will always slip through the net but, with every mistake, the companies’ systems will only become smarter, inspiring more faith in their players as a result.