In Ancient Greece, Apollo was heralded as the god of prophecy (and war). Priestesses who served him functioned as Oracles in the temple at Delphi. Thus, the infamous priestesses of Delphi became the source for many famous prophecies.
Contemporary society sits on the precipice of similar prophetic capabilities, through AI predictions. Artificial Intelligence technology may have the ability to predict future events sooner than we think.
The question becomes this: if we know our future, do we then become irrevocably bound to that future? More importantly, governments might decide to use AI to make future predictions. Will this affect policy decisions?
Asking whether AI predictions will still let society be free draws on the conversation philosophers have had for several millennia. That is, the question of free will.
Scared yet? Intrigued? Excited? Let’s talk about it.
AI Predictions and the Future of Society
Forecasting plays a major role in many industries already, including the following:
- The stock market
- Commerce and retail
- Fashion and culture trends
- Medicine and science
Artificial intelligence technology, as we can see, has universal applications across all parts of society. It has the ability to influence nearly aspect of our lives.
The processes of making predictions for forecasts and future trends is known as predictive analytics. Just like other automated data-driven technologies, predictive analytics continuously mines large data sets. Patterns found in past data get used to predict future actions or future events.
Analysts use algorithms to manually search for future predictions. They use a method called “regression analysis.” This describes tools that figure out the relationship between variables. The two types of regression analysis include:
- Single linear regression which measures two variables
- Multiple regression measuring three or more variables
AI technology develops tools to gather data in real time and it creates processes that automatically and quickly mine through vast amounts of unfiltered data to draw conclusions.
Machine Learning Prediction Mechanisms
Earlier this year, MIT led in the creation of a new technology that relies on a user-friendly interface to make future predictions. A lot of data-based AI predictions require experts to go through lengthy processes and input advanced machine-learning algorithms.
This new tool creates an alternative way for the average person to predict future trends. Researchers created what they call a “time series predict database.” This also can be called tspDB.
The interface runs on an already established time-series database. The database gets powered by a time-series-prediction algorithm. It processes predictions based on multiple variables.
When you collect data and store it in chronological order, you get time-series analysis. At MIT, they figured out how to run these multi-step operations with a simple model.
All users need to do is input queries which get immediately processed in less than a millisecond.
Predicting the Future
Making predictions about the future in itself does not set future events in stone. Rather, they attempt to find what might possibly happen in a potential future.
On its own, data analysis does not automatically lead to future predictions, because that relies on subjective reasons. So far, AI with all its capabilities does not do this. Computers and algorithms are not yet capable of causal reasoning.
This means inferring how cause relates to effects. Like the physics cliché, what comes up must come down. These types of predictions require the subjectivity of the human mind.
However, researchers continue to develop new algorithms that can perform the same functions as causal reasoning in order to produce future forecasts effectively. The Imperial College of London has done studies relying on Einstein’s light cones to model forecasting methods.
Light cones mathematically describe the confines of cause and effect in space time. It uses the same operations to measure predictions of how long it takes light to travel across distances.
These calculations serve as a data model for future predictions. It limits the number of frames that machine-learning algorithms can produce. AI-powered machines get trained to predict future frames based on a limited set of similar frames presented to it.
You can see the results of this research which has been published.
Societal Implications of Artificial Intelligence
We can already see many of the potential benefits that machine-learning predictive analytics can lead to. Doctors can use it to predict future disease diagnoses. Financial analysts and business people can track potential future cash flows based on predictive modeling.
AI also gets used to find insights about customer behaviors. Thus, AI predictions might become a tool to predict future human actions. Does this mean society will still be free?
Naturally, people will always be free to make their own choices. However, if futures forecasting tools calculate a limited set of potential future paths, this means decisions will be made based on these limited models.
Free will assumes we have a choice among an infinite set of possible futures. However, fundamentally we know that’s not true. Limitations based on systems of power like class, race, gender create altered courses that some populations get bound too.
Forecasting trends by nature uses historical data. If the future gets determined by the past, then do we choose our future? As mentioned earlier, philosophers have searched for answers to this very question since the time of the Ancient Greeks.
Imagine if your already limited options became pre-determined by machine-made calculations. Can a society be free if it lets computers determine a finite set of possible futures where the people in power only act on those models?
Forecasting Future Behavioral Paths and Trends
Just as in the very nature of ai predictions, we can only make guesses about what might happen in a possible future. Considering the volatility of human actions, especially as a group, no answers can ever be definitive.
An idealist would only see the good it could cause, but that assumes people (and the machines they design) will always act in good faith.
Because the future remains uncertain, the potential to make final decisions in present-times based on possibilities, instead of leaving it to fate, can limit society. At least, in theory.
Check out our site to read more about trends in technology and society.