# The Time is Now for Quantum

Although quantum computers may sound like the distant future, the age of quantum has arrived. Let’s summarize why quantum computers are special and see how this translates into a new industry with many applications.

You may have heard that quantum means “small particles” and we know this is misleading. Quantum effects can occur over large scales and long distances.

It is more helpful to think of quantum as **a new field of information science** — one that is particularly useful to compute “how complex possibilities will resolve into outcomes”.

### New Vocabulary — Superposition and Entanglement

There are two main properties that enable quantum computing, and here is a paragraph on each to get us started.

The first is **superposition**. Where a classical bit is always 0 or 1, a quantum bit — or qubit — starts in both states at the same time (!) and stays that way until measured. Algorithmically this is shown as a vector of a probability and a direction. What does it mean in English? That unlike normal bits that must always be “yes” or “no” the qubits can also mean “maybe”.

The second is **entanglement**. In a quantum computer, the qubits are entangled. This means that the state of each qubit affects the state of all other qubits, and vice versa, all in parallel and at the same time. **No qubit is an island.** So quantum computers can model situations where everything is interrelated. Quantum computers easily assess complex “what-ifs”.

### What Can We Uniquely Do With This?

Quantum computers are uniquely able to **model emergent behavior**. For example, for a classical computer to figure out how 40 people in a room who can communicate and affect each other’s decision will each ultimately decide to buy or sell and thus determine a market-clearing stock price, the number of interactions would be far too high to model exhaustively. There are 2⁴⁰ possibilities of buy and sell — more possibilities than there are stars in our galaxy. With a classical computer we normally have to simplify the complexity, e.g. we might model the trend of market averages.That would miss higher-order dynamics such as a run on the market.

Inter-related factors affect decisions all the time! Consider this famous scene in Princess Bride. Vizzini is confronted by the Man in Black and must decide which cup contains poison. He attempts to use his strength in classical logic and is quickly overwhelmed by the possibilities.

MAN IN BLACK:All right: where is the poison? The battle of wits has begun. It ends when you decide and we both drink, and find out who is right and who is dead.

VIZZINI:But it’s so simple. All I have to do is divine from what I know of you. Are you the sort of man who would put the poison into his own goblet, or his enemy’s? [pauses to study the MAN IN BLACK] Now, a clever man would put the poison into his own goblet, because he would know that only a great fool would reach for what he was given. I’m not a great fool, so I can clearly not choose the wine in front of you. But you must have known I was not a great fool; you would have counted on it, so I can clearly not choose the wine in front of me.

MAN IN BLACK:You’ve made your decision then?

VIZZINI:Not remotely!

### The Race to Develop A Quantum Processor

The hunt for a general purpose quantum computer traces back to comments by Richard P. Feynman at a conference on using computers to simulate physics organized by MIT and IBM here in Massachusetts. Feynman showed that a classical computer can not accurately simulate the real world because it cannot simulate quantum physics. He proposed that we instead create a new computer that operates by quantum forces, then load in the inputs and see what it does.

This is a really hard engineering problem. Classical transistors rely on semiconductors — materials that are not quite conductors and not quite insulators. What could quantum computers use? Brian Josephson earlier had been awarded a Nobel prize for creating the “Josephson junction” in which two layers of superconducting metals are separated by a thin nonconducting layer, achieving a delicate balance poised on the edge of conducting. In the late 1990s, IBM and others used microwave resonators to detect quantum effects in the Josephson junctions and built simple working quantum computers.

Since then, enormous progress has been made in improving these junctions and in pursuing alternate approaches. Other devices work with ions, others photons, and others operate at a near-quantum scale using whole atoms or oscillating drums.

### Faster Than Moore’s Law

Even in 2016, quantum computers were quite simple. IBM announced a model running on 5 qubits. This year, IBM simulated 49 qubits; meanwhile Google announced their 9-qubit processor can be scaled to 72 qubits with some noise, which they believe could be as good or better as a 49-qubit quantum processor at reduced noise. Why is 49 qubits a big change? While a 5-qubit computer can model 2⁵ possibilities; a 49-qubit computer could model 2⁴⁹ possibilities! Concepts for machines of 100 or even 200 qubits are on the drawing boards.

The hardware improvements in recent years are not just in the number of junctions but also in the quality — reducing the error of measurement — and in the speed — calculating faster and sustaining the quantum effect coherently for longer to allow for more operations. And these improvements compound each other. As a result, the usefulness of quantum computing is advancing considerably faster than Moore’s Law.

Next we need software. Quantum hardware is programmed differently from classical computers. Classical computers have boolean logic at the core. Quantum computers have all the bits connected so they must be programmed with matrix operations at the core.

New algorithms were invented recently that translate quantum capabilities into practical solutions while also correcting for the fairly significant error rates of current machines. Industry experts have declared that we are entering an era of Noisy Intermediate-Scale Quantum Technology (NISQ) in which the latest NISQ algorithms are expected to do useful work on imperfect hardware.

### A Key Moment

Are quantum machines ready for prime time? Almost yes. Classical computers can still calculate by brute force faster than today’s quantum computers can determine their results. However as we move to higher qubit counts and better qubits, and as quantum algorithms get smarter, we soon will achieve a moment of **quantum superiority** — where the quantum computer is faster for a given problem. Given the pace of progress, some feel the event is just 3–4 years away, others say perhaps 1–2 years.

An additional tailwind to drive this home is coming to the USA, where quantum research was just recognized as a national priority and grant programs at the DOE are expanding.

Implication? Our start-up community needs to prepare for exponential change. Once quantum computing becomes obvious, it could be moving too quickly for a new company to jump on from a standing start. Since it takes several years to build a start-up of any serious capability, that means **forward-thinking investors should prepare to deploy seed capital as early as now**.

### Where to Start and Back New Companies

How can founders ride the coming wave of quantum computing? A full stack is emerging.

At the **infrastructure layer**, there are plenty of opportunities for better and more innovative hardware, for developer tools, and for better algorithms.

Just as AWS helped software start-ups by reducing the investment cost of entry, the quantum computing field also supports a cloud-based service model. IBM, Google and others already allow software developers to run code on their quantum hardware in the cloud. Rigetti, a quantum hardware startup backed by a long list of VCs, announced a full stack cloud service that provides both classic and quantum computers on one platform for ease of use. Chinese firms are just as active — Huawei just announced last week that their 42 qubits support has already been implemented as a cloud service.

Pillar VC invested in Zapata Computing to develop next-generation algorithms and offer a layer of cross-platform software that runs on these clouds. Zapata Computing has already announced close relationships with IBM and with Google and research results with Rigetti. Working with large enterprises to design custom solutions for their quantum-scale problems, Zapata is accumulating a general library of practical approaches — including smart intermixing of classical and quantum techniques — that will become a future **middle layer** to simplify life for developers.

At the **application layer**, inter-related systems are all around us, which means that quantum computers could by applied in many ways:

Today, training multi-layer neural networks is slowed by limited compute speeds, even with special hardware. A quantum computer may be able to train a more complicated model more quickly. The fields of **artificial intelligence and machine learning **are likely to jump significantly in sophistication when quantum computers arrive.

Market prices are the result of emergent behaviors. Quantum computers may predict **financial markets **with greater foresight.

How two molecules interact is influenced by a complex interplay of electron attraction and repulsion as well as physical characteristics that can be dynamic such as rotation and flex. Quantum computers could vastly improve design of **new chemistries **like batteries, fuels, fertilizers, refineries and catalysts. These techniques could be enormously helpful in **pharmaceuticals**.Whether a small molecule will affect a particular protein is determined by a simultaneous interaction of many dozens of binding sites. (The recently retired Global CIO of Merck, technology vertan Clark Golestani, recently joined the Board of Zapata Computing.)

There are a long list computer science problems may be solved better with quantum **optimization**. **Scheduling** factory work and air flights, **searching **websites, **buying **advertisements, **ranking** user preferences for playlists, **predicting** fashion trends, **personalizing** health interventions, **placing** retail outlets, **modeling** the weather, and **routing **directions are examples.

Finally, in the field of cybersecurity, today’s encryption is enabled by the use of big number factoring — a math problem that is too hard for classical computers. Quantum computers though, could factor in seconds what would take classical computers a billion years to solve. We are going to need **a new basis of encryption**. That could be a new type of math problem, or, it could mean using the entanglement feature of quantum mechanics to distribute one-time-use “quantum keys” across long distances.

Each of the areas above could support a pure-play start-up to develop and then widely provide complete solutions for vertical markets.

### Boston Can Play a Key Role in Quantum Industry

Looking out over the next few years, quantum computers will drive a substantial new wave of innovation. As we meet a growing list of quantum enthusiasts here in Boston, we observe that Boston has an opportunity to lead this revolution. We have the concentrations of quantum hardware talent, computer science talent, enterprise software talent, AI/ML talent and of course cybersecurity talent **available here in Boston**.

If that includes you, now is the time to come up to speed on quantum computing! Once you see how you want to apply this fascinating new capability, we welcome a discussion with you.