Opinion The UK’s national broadcaster, the BBC, its R&D team and all of its 15 million century-old archives are part of a new student consortium QNLP, Quantum Natural Language Processing, with the ultimate goal of automating the extracting meaning from the babbling of humanity.
“The most incomprehensible thing about the universe is that it is comprehensible,” is one of those rare Einstein quotes that Einstein actually said. We don’t know what he could have said about Monthy Python’s flying circus as he died 14 years before his first transmission. But it’s fascinating to wonder what he, as one of the founders of quantum physics, might have thought about the idea that quantum computing tells why the universe is understandable in the first place.
The consortium, announced on November 25, is receiving funding from the Royal Academy of Engineering and will build on the quantum mechanics and linguistics work of Professor Bob Coecke, chief scientist at UK-based Quantinuum; Professor Stephen Clark, Head of AI at Cambridge Quantum; and Professor Mehrnoosh Sadrzadeh from the Department of Computing at University College London. Two geeks in a garage, that’s not it.
Long-time followers of quantum computing news will know that every QC story exists primarily in the future: the technology is more promising than the product. It is limited by the current state of the art, noisy intermediate scale quantum or NISQ. Current systems are too noisy and too small to be useful. Much of today’s QC research is about developing techniques and algorithms that will become the best in the world, once we move out of NISQ and into large-scale fault-tolerant systems. QNLP is no different.
What makes it interesting is where it comes from. The professors and their teams have 15 years of research in language analysis to their credit. One of the results is the beautiful DISCOCAT (DIStributional COmpositional CATegorical) framework, which creates a data set from groups of sentences that can be analyzed on a quantum system. The intrinsically interesting part of this is that DISCOCAT produces a tensor network that corresponds very closely to the natural workings of quantum logic. The project says it’s inherently a good fit for quantum mechanics. But very few standard computing tasks are, so why would it apply to meaning encoded in language?
The answer, say the researchers, is category theory. It’s a mathematical approach to systems analysis, first mooted in the mid-twentieth century, which says you can learn a lot about a system by ignoring the internal details of each component and focusing on how they interact. By providing a map of behaviors, category theory can reveal patterns that cannot be easily derived by trying to break down the individual components – making it a very good fit, for example, quantum mechanics. Categorical quantum mechanics is a recent field of study that focuses on patterns and processes at quantum levels, making it a good choice for quantum logic, among others.
Category theory is also a good complement to linguistic analysis, producing meaning maps that include information about the relationships between grammar and semiotics – the structure of meaning encoding. It’s both extremely useful and, for AI researchers and philosophers of mind, a very tempting avenue for conceptual exploration.
Most important, however, is category theory’s ability to find similar patterns in seemingly disparate systems. It’s basically how math and physics advance, using knowledge of one system to better understand another. What the consortium researchers say is that the quantum nature of their linguistic analysis comes from the fact that it works on models similar to quantum mechanics. Therefore, QC will be incredibly good at language – when it works.
This connection has been theoretically known for some time, but limited to classical computer simulations. Now, reality is proven to be ready to conform to theory, with recent experiments beginning to ask small questions from small sets of sentences on IBM’s Quantum Experience platform. These involved only a few tests, one to ask which of a hundred sentences was about food and which was about computers, and one to tear off noun phrases. Classic computer simulations then run alongside quantum tests to show what you could gain when fault-tolerant large-scale systems arrive.
In that regard, it’s as good as QC gets. But in the sense that a fundamental tool of mathematics and information science makes explicit connections to the deep structure of language and the workings of quantum mechanics, it’s a very intriguing clue to how quantum computing is as interesting for philosophers of cognition as it is for physicists, companies and computer scientists. Language is a function, perhaps the defining function, of how we classify ourselves as intelligent, and language processing is an intrinsic and unique part of human cognition and human society. Finding it obeying rules that other physical systems exhibit does not mean that consciousness is more quantum than any other macroclassical system; nature replicates patterns at all scales, after all.
But it may help explain how we can find so much of physics comprehensible; it follows patterns that we are set to exploit. Finding a potential answer to something that baffled Einstein is no small feat. And who knows, when a future post-NISQ AI has digested all the BBC production, we might even be able to ask him not just what the Parrot Sketch means, but what daytime television is for. Maybe that’s a philosophical question too far. ®
#Creepy #tangle #revealed #Quantum #BBC