Inspiration
The hardest tool to truly master is problem-solving. We see the thinking capabilities of machine learning increasing, but equally so our thinking capabilities correlate. The inspiration came from the thought experiments when paradoxes arise that take thoughts to a different space. To show that thinking is more than simply one or its opposite.
Goal
The goal of this experiment is not to yield specific results or determine a truth in results, but instead provoke thought for those that examine it. It is meant to suggest the thinking capabilities of ourselves and how it can lead to more effective AIs. The purpose being to create logic within a world full of quantum ideas. To simply demonstrate thinking is not binary (1/0, or True and False), but instead quantized (infinitely many solutions).
Try it out
https://colab.research.google.com/drive/1X4GbbY0epW76GPC3RCGVzr3VDDm2zRrs?usp=sharing
What it does
(Save reading this portion for after taking the test to understand its functions)
This test gives portions of a point based on the most accurate demonstration of quantum (more possibilities than 1 and 0), not to score “correctness” or “intelligence”, it is instead meant to be open for interpretation as it has no logical “more right” score. In the first half of the test there are 10 questions. These questions are directed towards choosing the “most correct” answer on a scale of one side being -10 to the other side being +10. Points were given based on how close the user put an answer to 0, since 0 was a show that neither of the two options were more correct than the other. The points were given in fractions of the distance. These were paradoxical questions where the answer cannot be True or False.
The second portion of the test gives the opportunity for open answers. The test didn’t take into consideration the answer it was given, but instead it measured how long the user took to answer the question. When the user stays in thought they consider both options to be both true and false at the same time, just flipping to determine which one follows a better path of logic and optimization. Once an answer is inputted the point is given based on if the user spent a specific amount of time on the question, fractions were given as it was a percentage of the time.
Thinking is more than binary, the state follows some form of true or false, but instead through infinite interpretation. This is where time constraints can be given for machines to think through infinity and determine the best solution within that given moment.
How we built it
By understanding the way computers face logic, we can create more efficient ways to debug programs. A code when given a paradoxical statement runs into an infinite loop of problem solving. It never comes up with a solution because it is only thinking in binary as the answer having to be True or False. If instead given a time constraint, The code could instead come up with the optimal solution within the given time.
The code itself was created through python and was done only by receiving inputs and doing calculations for the points. The points were given based on a fraction of scored versus the total possible. If the user scores under <50% they are closer to binary thinking and if they score >50%, then they score closer to quantum thinking. If the user scores 50% then they fall right in the middle!
This percentage is meant to only evoke thought and if you took this test again you will most likely score differently when taking different factors into consideration while coming up with an answer.
Challenges we ran into
It is difficult to create a test that can truly accurately describe something that is paradoxical in itself. An example of an issue that arises is the time factor for the test in determining what the optimal time is to receive a full point. How does one determine what the most correct time is? Another example of an issue that could be how the percentage at the end is determined.
The original idea was to follow a logic regression model and plot the average score as the middle point of the graph where the user is at exactly 50%. This requires more testing to get a better average score as well as to find what the optimal timing is for the given moment. Creating the best version of this will help demonstrate how binary computers might look at information versus a quantum computer given the correct abilities.
Accomplishments that we're proud of
As creating a test such as this one might not make complete sense to even us, the primary objective is to put solutions to unsolvable problems. It helps us to look at the world from a more positive perspective as it has infinite ways to make it better! The fun is trying to push light into the unknown things such as unsolvable mysteries such as paradoxes.
. I appreciate if you have read this far :)
What we learned
Lookin through space
and time from different perspectives gives us different insights and helps us understand the universe and our place in it better. By combining different perspectives, we can get a more complete and nuanced view of reality. /ChatGPT
What's next for Binary Thinking vs Quantum Thinking
As this is merely only a thought experiment, it is truly to describe the capabilities of thought and quantum computing. If an AI could learn to be more centered in a paradoxical state, it could come up with better solutions. Instead of leaving it in an infinite True or False state, it can instead think individually down paths of logic to find an optimal solution in the moment.
Moving forward in time gives a state of infinite possibility. Problems may have no solution, but working to find one helps to make things just a bit better. The more we learn about AI and ourselves, the better we can find ways to solve these unsolvable questions.
Log in or sign up for Devpost to join the conversation.