AI Lexicon — Q

Quantum machine learning (QML)

Today's digital computers are fast, much faster — as we are often reminded — than the computers that got humans to the moon in 1969. Supercomputers are incredibly fast — their processors use what's called "brute force" to work through many variations of a thing. But even they can't handle the most complex problems, such as simulating and analyzing every possible permutation of a molecule's structure of function. Tomorrow's quantum computers, on the other hand, may just be able to do it.

Digital computers are built on a binary code of zeros and ones — each unit of information can only be one or the other so-called "state" — a zero or a one. Complex information can be a series of zeros and ones, but it's still only a string of zeros and ones.

Quantum computing is built on qubits (quantum bits), a far more complex unit of information. A qubit can represent a one, a zero, or a "superposition".

IBM describes the superposition as representing every possible configuration of a qubit: "Quantum algorithms [create] multidimensional computational spaces. This turns out to be a much more efficient way of solving complex problems like chemical simulations."

A Microsoft Azure resource says this: "It would take a classical computer millions of years to find the prime factors of a 2,048-bit number. Qubits could perform the calculation in just minutes."

Germany's Fraunhofer Institute for Intelligent Analysis and Information Systems, meanwhile, says quantum computing could be "the springboard for a huge leap forward in artificial intelligence."

It's all about the amount and complexity of data that researchers think we will want to process in future. AI is often used to combine different data and calculate the best (or "optimized") solutions for given problems. As a result of the increasing amounts of data in the world, some researchers say conventional computers, even supercomputers, will be too slow to handle it all and, as IBM puts it, "stall." Quantum computers may even be able to cope with incomplete or corrupt data.

But the Foreign Policy publication has written that with its projected ability to do so much more than classical computing, "quantum computing is even more dangerous than artificial intelligence." (za/fs)

 

Sources:

Towards Quantum Artificial Intelligence (Thomas Gabor, QAR-Lab, LMU Munich) https://www.mobile.ifi.lmu.de/wp-content/uploads/team/thomas-gabor/towards-quantum-ai.pdf (accessed October 18, 2023)

Quantum AI (Fraunhofer) https://www.fraunhofer.de/en/research/current-research/quantum-technologies/quantum-ai.html (accessed October 18, 2023)

Zuse's Z3 computer and the digital revolution (DW, Abbany) https://www.dw.com/en/konrad-zuse-and-the-digital-revolution-he-started-with-the-z3-computer-75-years-ago/a-19249238 (accessed October 18, 2023)

What is a qubit? (Microsoft Azure) https://azure.microsoft.com/en-gb/resources/cloud-computing-dictionary/what-is-a-qubit (accessed October 18, 2023)

Why quantum computing is even more dangerous than artificial intelligence (Foreign Policy) https://foreignpolicy.com/2022/08/21/quantum-computing-artificial-intelligence-ai-technology-regulation/ (accessed October 18, 2023)

What is quantum computing? (IBM) https://www.ibm.com/topics/quantum-computing (accessed May 7, 2024)

Read the rest of DW's AI Lexicon:

A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z

We're keen to hear your feedback. Suggest an entry by sending us a comment. And let us know if you feel we have missed something, got it wrong, and tell us whether our AI Lexicon has helped you understand the technology better.

Written and edited by: Zulfikar Abbany (za), Fred Schwaller (fs)