After PokerStars hire a couple of AI experts, Lee Davy, goes off on one about the end of the world.
Imagine handing the keys to your Gmail account to a stranger. Would you do that? How about giving them to an artificial intelligent (AI) algorithm?
Between 2008 & 2012 a study group at the University of Cambridge asked 86,220 Facebook users to complete a 100-page personality questionnaire to determine how well they knew a work colleague, friend, family member or spouse.
They set an AI the same task by analysing the number of “Likes’ a person had made on their Facebook feeds. The AI only had to examine ten Likes to out-predict work colleagues. 70 likes to do better than friends, 150 Likes to perform better than family members, and 300 Likes before
the machine knows you better than your spouse.
If an AI can become that familiar with you via your Facebook Likes, imagine how intimate the AI would become with your personality traits if it read your emails?
You would have the greatest personal assistant in the history of the world. One that would be able to tell you which Tinder dates you are more likely to end up at the altar with, what present to buy your father at Christmas, and how to sift through The Stars Group job vacancy list.
Seriously, have you seen how big The Stars Group vacancy list is?
The Rise of the Machines
Unless Kim and Donald manage to wreak so much havoc on the world that we lose all computer systems, and the knowledge of how to rebuild them; it’s safe to say that it’s a matter of ‘when’ we create human-level artificial intelligence, and not ‘if?’
Elon Musk, a man who has spent many sleepless nights thinking about existential threats and what his billions can do to prevent them, believes the single biggest concern for humanity is not a nuclear war or climate change – it’s the rise of AI.
AI will be used to do tremendous good in the world, but we must remember that there exists a group of people more than willing to wipe out everyone on the planet, because they believe in doing so, gives them a passage to an otherworldly paradise.
Imagine a powerful AI in the hands of people willing to fly planes into buildings.
And consider this.
I woke up, this morning, and saw a spider crawling on a cuddly panda lying next to my one-year-old daughter’s head. I am in the US. I don’t know if it’s poisonous. I freak out, pick the panda up, and shake the spider into the toilet, and flush.
Why is it so foolish to think that an AI wouldn’t want to flush us down the toilet?
Wouldn’t we be a threat to AI in the same way a spider would be a threat to my daughter?
All of this means that we all have to start thinking about the effect that AI will have on our future, and cross our fingers that bright people brighter are busy working on a way of mitigating AI risks.
One such non-profit is the Machine Intelligence Research Unit (MIRI), and recently PokerStars donated a significant portion of their Spring Championship and World Championship of Poker (SCOOP & WCOOP) $25,000 High Roller fees to MIRI via Raising for Effective Giving (REG).
REG is a non-profit created by poker players to help reduce suffering in the world, and one of their interests is the rise in AI. Two REG founders, Liv Boeree and Igor Kurganov, are employees of PokerStars, which must make for some exciting team meetings.
AI and Poker
AI already dominates the world of poker.
During a recent interview with the 888Poker Ambassador, Dominik Nitsche, as part of my Poker Routines column, I asked the German to describe his process of review:
“Everyone knows there are solvers out there these days. They are better than any human can be. I try to learn a lot from software like this as I find most of my opponents tend to try and copy a solver style.”
Most online poker rooms have introduced mitigation to ban Heads-Up Displays (HUDs) and seating scripts; AI designed to give the superior players, even more, superiority than the weaker players in the game.
And then you have the bots, AI created to infiltrate online poker games so it can win money playing against weaker human opponents.
So far, the online poker operators have been trying to mitigate the rise of the bots by amping up security. But this is a fools game. Nobody who takes the threat of AI seriously is looking at creating a security system to prevent them from flushing us down the toilet. The smartest people in the world, people like Elon Musk, are working on ways of turning humans into cyborgs so we can match them or overcome them should they decide to don a pair of boxing gloves and go toe to toe.
And if you believe turning humans into cyborgs is a little too Blade Runner for you. We are cyborgs. Our coliseum is virtual, and our trident-wielding gladiators are spectacled brainiacs. Instead of trying to beat the bots (some pros now call Libratus, and not Phil Ivey, the best poker player in the world), we have to join the bots.
In a recent interview with Virtue Poker’s Ryan Gittleson, he confirmed that the peer-to-peer decentralised poker platform would include the ability for bot operators to take on other bot operators in combat. Gittleson will also allow bot v human matches. It’s the most transparent and most sane response to AI that I have witnessed in the poker industry, and it appears Boeree and Kurganov are also waking up the execs at PokerStars.
In an article written for PokerNews, the former PokerStars employee, Michael Josem, drew attention to that huge vacancy list I was talking about and two positions in particular.
Graduate Poker AI Research Engineer (London, Dublin)
Poker AI Research Engineer (London, Dublin)
Josem notes that the positions report to the Poker Innovation Team and not Game Integrity, suggesting that PokerStars may be considering how to embrace AI to improve their product, rather than trying to protect their customers from a seemingly unstoppable threat.
You can read Josem’s thoughts here. In a nutshell, he believes PokerStars are looking at using AI to create more products, both live and online.
PokerStars has already shown with the release of Power Up that it’s ready to pivot to keep up with the advancement in trend and technology. One of the areas that Virtue Poker have not considered is human/bot teams competing against human/bot, human/human, bot/bot teams.
The hybridisation of the game is what happened in chess after Deep Blue beat Garry Kasparov all those years ago. And the world’s best chess player is a bot/human hybrid.
Is this a route that PokerStars could turn to?
If they do, I believe it will be a temporary measure.
Management structures are much like political structures. There aren’t too many people looking to the future. Donald Trump is thinking about the next four years. The Stars Group execs are looking at this budget year.
Thank God we have Elon.
How long before we look at AI as our God?
When you contemplate the seriousness of this situation, it’s ridiculous to think there will even be such a thing as online poker in the future.
And what do I mean by ‘future?’
During my recent visit to the Effective Altruism Global Summit in San Francisco, the talk around the dinner table was human-level AI would be born somewhere between the next 10-50 years.
PokerStars, you have plenty of time before we are flushed down the toilet.
If you want to learn more about the rise of machines in the gaming world then check out this neat infographic.