Quantum Computing

But it wouldn't be measured but acted on instead.

It is very likely that all electronic/digital data created at the very least since 9/11 or shortly after has been captured or acquired. All of it.

The vast majority of historical data prior to the digital age has also been converted from paper, audio, photographic to electronic data. This would include things like ancient manuscripts and books covering all of prior human intellegence.

Electronic/Digital Data: Any data stored in static form (files, docs, PDF's etc.) on storage devices along with any data transmitted (bytes) not intended necessarily to be stored.

Once all known human data (intellegence) has been digitized (binary) into a single language it can be analyzed.

If an advanced AI program is written (I'm sure it has been) it can analyze all human intellegence, learn from our prior actions and habbits and develop a very accurate understanding of our behavior and how our actions affect our behavior and what outcomes they have produced good or bad. This program would also have an almost perfect understanding of our limits (governing structures, laws, morality, immorality, psychology etc.).

How does one test AI program for accuracy? Simple. Provide it with a scenario (beginning of WW2 for example) and see if it arrives at the same conclusion humans did (defeat of Nazi Germany) with a great degree of accuracy. Notice I said the testing was simple. The hard part would be the refinement and adjustments to get it to the point of accuracy.

Once the AI program has been run through enough historical scenarios and arrived at the same conclusions humans came to with enough accuracy and frequency than one may be able to safely assume it knows enough about human behavior in order to accurately PREDICT human behavior in the future. Predicting is NOT doing. In other words the AI program would not initiate anything on its own. It simply waits for input (data) and provides its best analysis of outcome based on factors or data that has been provided to it. At which time human can make decisions based on the AI's prediction. The give and take between human input and AI output could go on until infinity based on human action NOT on AI action. In other words AI is simply telling human what is likely going to happen given the scenario human just gave it. If human sees that his actions brought on nuclear war than human changes his behavior and re-runs the exercise. So human tells AI what it move it plans to do on the chess board, AI tells human what his opponent will counter with with an astronomical degree of accuracy. If human sees the counter will be bad for himself he simply doesn't make the move and changes strategy. Re-runs the chess moves and doesn't actually physically move the chess piece until he is happy that a counter move will benefit him.

Ok back to omniscience.

If AI has all prior human knowledge and history, knows our behavior and psychology it can predict our opponents moves given our moves. If AI has the power of super computing, is connected to EVERY data stream in REAL TIME as its happening (email, phone, text, news, covert intel etc) than it could in theory give its owner the ability to make moves and react in near real time to opponents counter moves in a way that benefits the owner and traps or places his opponent in a very bad position.

If you want this AI to step it up a notch, give it your opponents stream of thought. In other words access to your opponents emails, texts, live phone calls and all communications. Humans often seek advice and or broadcast our thoughts and intentions to loved ones and confidants before we even take those actions. AI would provide a set of possible predictor variables to owner before the opponents move is even made. You can now see how the owner could be well ahead of his opponent.

A Super Computer can easily do all of what I mentioned above. A Quantum Computer can model just about every human on earth in real time and make moves and counter moves based on earth's aggregate population actions and behavior. It simply needs real time data. I think we all know if there is one thing more plentiful than water it is real time data. We're pushing it out without a care in the world through every copper wire, fiber optic cable and frequency available. After all this AI program couldn't possibly exist...right?