With Mueller having found the witch that Trump claims he’s hunting for (“A total WITCH HUNT”), the special counsel investigation has hopscotched a square closer to its end. March 28, the New York Times reported that Mueller linked top Trump campaign aide Rick Gates to a former member of Russian military intelligence.
On the other hand, Trump doesn’t like witch-chasers as indicated by leaks that report Trump’s desires to execute an Apprentice “I fire thee” slam at Mueller.
Over the last week, Trump fired a volley of frenzied tweets blasting comrade Mueller, and letting loose on the No Collusion, the Dossier and all Trump’s favorite items.
“The President,” said CNN, “wants this whole thing over with. Now.”
AI: Trump or Mueller?
Over the last few years, AI proved its ability to uncannily predict results on a large range of issues. For instance, AI can predict your death with 90% accuracy. It can predict whether you will acquire psychosis or mental disorders, including schizophrenia. Artificial Intelligence can predict whether you or your children will lapse into criminal behavior. In 2017, AI accurately predicted first that Trump would be TIME’s Person of the Year, followed by the accurate prediction that the #MeToo campaign would splash TIME’s cover in 2017. Artificial Intelligence predicted Putin’s candidacy (frankly, a no brainer) and accurately predicted the 90th Academy Awards, with a stunning 94% probability. The Swarm AI, also, perfectly forecast Trump’s 100-day approval rating before he even took office.
How does AI predict Mueller’s demise?
In 2017, the researchers at Unanimous AI fed the following questions to their system.
- They asked their AI to to forecast the percentage of Americans who approve of Mueller’s handling of the investigation. The program produced an accurate response, despite the researchers not having provided it with this information beforehand. Ninety-two percent of US voters approved of Mueller’s handling of the investigation.
- The researchers asked the Swarm whether Trump would emerge from the Mueller investigation unscathed. The swarm was divided in its response, with the option leaning to ‘Yes’.The indictments, it said, were 76% “very damaging” to the President.
- Between calls for Mueller’s removal from the right and Trump’s removal from the left, researchers wanted to know who would be more likely to leave office by the end of 2018? The swarm gave this a 77% probability.
The upshot: Mueller will be more likely to leave his current role than the President – although how Mueller leaves and why will have a major impact on Trump’s presidency.
Other questions included:
- When will the investigation wrap up? Swarm: Before the 2020 General Election (57%)
- Odds that Trump will pardon someone related to the investigation? Swarm: 20%
- How likely is Mueller to leave by the end of 2018? Swarm: Highly likely (82%).
According to scientists like Colleen Farrelly, data scientist at Kaplan University, AI works best on non-stochastic events, where there is minimal to no interference of randomness. Chess is a good example. I can move a pawn from A2 to A3 and know it will always work. There is no uncertainty. On the other hand, take a game of dice and I can never be certain that a specific number will be rolled.
When it comes to AI, predictions based on hard or physical science like those of the weather are more likely to be accurate, since they’re predictable. Hence, you can bet on aspects like elderly people dying around a certain age range and know almost certainly that you’ll win. As I type in the first few characters of a word in a Google query, Google may present me with a list of potential words. Its machine learning program has used a long history of previous words to almost always accurately predict what I want to say next. These type of predictions are based on straightforward physical data that contain little randomness.
Humans, on the other hand, are messy. They’re moody, they’re irrational. You never know from one day to the next what will occur in,say, politics (especially, with the Trump Administration!). Therefore, AI is more likely to falter when it comes to social events.
As Farrelly says:
“I can design an algorithm to take in historical patient and 6-month service utilization data and spit out predictions about 6-month risk for service utilization for a healthcare system (have done that with 99.99% accuracy). However, making a prediction, like who will win the 2020 election is not possible (and probably won’t be possible in my lifetime, if ever). That’s the problem with dynamic systems that have stochasticity in them (like most real-world systems)–it’s essentially a chaos theory problem with volatile outcomes and sensitivity to initial conditions at multiple points of the process.”
To return to Mueller, I see plenty of possible volatile outcomes. For instance, Trump would have to fire the deputy attorney general, Rod Rosenstein who appointed Mueller. If Rosenstein refuses ( which he already said he does), Trump could fire Rosenstein and repeat the process through the line of succession, until Trump finds someone who will obey. Trump could defang the Russia probe by finding one of more of its strategies or techniques “so inappropriate or unwarranted under established departmental practices that it should not be pursued.” Democrats could gain control of one or both houses in the November midterm elections, making it more difficult for Trump to fire Mueller. Trump could die from his McDonald fries (unlikely). Kim Jong-un and he could lobby nuclear shards and that’s the end of the Russian investigation. More unlikely. Hopefully, if AI proves correct in this case, too, so all who care about the US welfare will take to the streets and protest. So, then Mueller would lose his job – and Trump would, too.