Towards the end of 2019 I finished a book AI concepts for business applications. The last chapter was entitled “The Future”. I wrote about quantum computing and a related version of deep learning: a “Quantum Walk Neural Network”.
In 1980 the idea of a quantum processing unit was proposed. Such a processing unit does not use the ones and zeros known to us. This “classic” mindset is the way weather think with a 1 for true and a 0 for false, and combinations – for example a “false positive”. Quantum computing is based on a “superposition” of states called “quantum bits” or “qubits” for short. But there is a huge difference between the way we do it think and the way of nature behaves.
In 1981 the late Caltech professor Richard Feynman (a Nobel Prize winner for his work on “quantum electrodynamics”) summed it up: “Nature is not classic, damn it, and if you want to do a simulation of nature, you’d better do it quantum mechanically, and with God it’s a wonderful problem because it doesn’t look that easy. “
Now quantum computing is starting to emerge. It started with hardware:
- In March 2017, IBM announced an open application programming interface (API) called IBM Q, where Q stands for Quantum.
- In December 2017, Microsoft announced an unbeatable preview version of a developer kit using a programming language called Q #.
- In January 2018, the world of neural networks, which includes a convolutional neural network (CNN), mainly for images, and a recurrent neural network (RNN), mainly for text, was expanded to include a quantum walk neural network (QWNN). The QWNN paper entitled “Quantum Walk Inspired Neural Networks for Graph-Structured Data” was written by Stefan Dernbach (then a PhD student at the College of Information and Computer Sciences at the University of Massachusetts); Arman Mohseni-Kabir (then PhD student in physics at UMass Amherst); Don Towsley (Dernbach’s PhD supervisor); and Siddarth Pal (a scientist at BBN Raytheon Technologies).
In their abstract they wrote: “A QWNN learns a quantum path on a graph in order to construct a diffusion operator that can be applied to a signal in a graph. We demonstrate the use of the network for Prediction tasks for graphically structured signals. “
Notice the phrase “predictive tasks”. This is what deep learning is known to do, that is, a model “for the label” (or a category or classification), once trained with tagged data, is able to identify images or text from a flood of inputs that the model has never seen before and still find the needles that match the model. Such models have come to be known as “prediction machines”.
- In March 2018, Google’s Quantum AI Lab announced a 72-qubit processor called the Bristlecone.
- On July 19, 2018, Google announced an open source framework called Cirq (where the C stands for cryogenic) and is planning a Bristlecone cloud.
- On January 8, 2019, IBM announced IBM Q System One as the first integrated quantum system for commercial use.
- On February 21, 2019, Google announced a cryogenic controller that only consumes two milliwatts of power.
- In May 2019, Microsoft announced that it would publish parts of its Quantum Developer Kit on GitHub as open source in the summer of 2019, including the Q # compiler and quantum simulators.
- On October 23, 2019, in a nature Paper proclaimed “quantum superiority” to Google. The paper was entitled “Quantum Supremacy using a Programmable Superconducting Processor”. As Google summarized the progress in the abstract:
A fundamental challenge is to build a high-fidelity processor that can execute quantum algorithms in an exponentially large computing space. Here we report on the use of a processor with programmable superconducting qubits2,3,4th,5,6th,7th To generate quantum states on 53 qubits, corresponding to a computational state space of dimension 253 (around 1016). Measurements from repeated experiments sample the resulting probability distribution, which we verify with classic simulations. Our Sycamore processor requires approx. 200 seconds Sampling an instance of a quantum circuit a million times – our benchmarks are currently showing that the equivalent task of a modern classic supercomputer would take roughly 10,000 years. “(Bold added.)
From all of this one can see that the field of quantum computing has finally made it onto the launch pad of an “emerging technology”.
Quantum computer patents
With this story we switch to patents. I previously presented bar charts for two new technologies: deep learning and blockchain. These graphics are based solely on a search in the patent database of the US Patent and Trademark Office (USPTO).
As before, I looked for a keyword or phrase in the damage field in the USPTO database. For the annual data, I searched the USPTO for “Quantum Computing” in the claims and for the issue date on an annual basis. The bar chart for “Quantum Computing” is surprisingly similar to the bar charts for deep learning and blockchain.
THE QUANTUM COMPUTING PATENT LAND RUSH
The total on November 16, 2021 was 322. Remember, the 2021 total is for a sub-year beginning November 16. Since there will be six more Tuesdays in 2021 (when new patents are announced), I forecast 150 or more by the end of 2021.
If you compare this bar chart to the charts for deep learning and blockchain, the conclusion is easy to see. We live in a time in which deep learning, blockchain and quantum computing are emerging quickly and almost simultaneously. These advances will create miracles that we cannot foresee now.
If readers know of any other candidate for a new technology, please let me know in the comments below.
is a lawyer with two engineering degrees – a BS in Engineering Systems from UCLA and an MS in Environmental Engineering Science from the California Institute of Technology. From 1975 to 2014, he practiced as a litigation specialist in California, representing plaintiffs and defendants in both federal and state courts. For the last 18 months of his career, he was of Counsel to Cotman IP, a patent law firm in Pasadena, CA. He is also an inventor named on eight US patents and the founder of Intraspexion, a Delaware LLC that owns patented software to implement “deep learning” in the context of “threats or risks of interest”.