Successfully used computers to predict the future

"Prophet" computers use quantum computing, transcending traditional computer barriers that use a binary computing system.

The team of scientists from Griffith University, Australia and Nanyang Technological University (NTU), Singapore has developed a quantum machine that accurately simulates future results.

This research is of great significance at the time that quantum mechanics and machine learning (artificial intelligence) is gradually interfering like this.

Prophetic ability is no longer a remote thing thanks to the power of computers.(Photo: Yourminds).

The team created the quantum optical information processor, using the quantum superposition principle to determine the possible orbits of a photon in the light beam, then align them in order of probability.

The computer "prophet" computer

People spend a significant portion of their time making decisions based on expected results or theory. For example, when we hear a timer that turns off the stove, we check the stove because the potential result of ignoring the alarm signal is that the food is burnt. When removing cooked food from the oven, we choose not to take the pan or tray with bare hands because it is anticipated that the hot object will burn the skin.

This type of decision making is very easy to implement, because people can rely on available data. But in the world of computers there are not only a few options. People have to deal with large-scale data problems, such as how to optimize random process predictions. These are systems with random variables, so it is impossible to know in advance the output that must be used to solve the problem.

The team succeeded in determining the possible orbits of a photon in the light beam, then placing them in order of probability.

"When we think about the future, we face a lot of possibilities. These capabilities increase exponentially when going into the future prediction. For example, when there are only two possibilities to choose each minute. In less than half an hour, there will be 14 million possible results, in less than a day, the results will exceed the number of atoms in the universe, " said Assistant Professor Mile Gu of NTU Singapore. , the experimentalist said.

Therefore, physicists and machine learning experts are studying quantum solutions for problems beyond the limits of classical computers (not enough memory or memory is not fast enough).

The era of quantum computers

Machine learning researchers often focus their efforts on combining powerful computer hardware (often using thousands of GPUs or CPUs at the same time to train systems to run predictive algorithms), and develop algorithms more powerful to solve problems with millions of variables such as an artificial neural network.

Quantum computing allows us to solve more variables with less power. Scientists believe that quantum systems will have an advantage over the classical system. They will surpass the limits of the binary computational system, which has only 1 and 0, because there are practically numerous states between 1 and 0. In addition, quantum computers also help to improve computational capacity. and memory optimization.

Quantum computers allow us to solve more variables with less power.) Photo: Nikkey).

The group's work has a profound meaning for the field of AI, especially deep learning. It can be used to significantly improve the training time of artificial neural networks for natural language processing, image recognition and other common AI applications.

According to the researchers, predicting the successful photon trajectory has confirmed the advantages of quantum computers. This technique is capable of reducing the necessary capacity of computer memory in simulating random processes, while providing tools for quantum artificial intelligence.

Currently the test device can only simulate 16 future possibilities. But in theory, algorithms can expand to infinite.

1. 6 applications that change the world of quantum computer technology
2. For the first time, quantum computers have been proven to be more powerful than classical computers