Embedding Ethics into Quantum-as-a-Service: Q&AEmbedding Ethics into Quantum-as-a-Service: Q&A

Quantum computing is set to transform many industries, including aerospace, health care, logistics and financial services, and power the next generation of AI.

Most organizations access, or plan to access, quantum computing resources on genuine quantum hardware or GPU-based emulation via the cloud using a subscription resource known as quantum-as-a-service (QaaS).

One provider, Dynex, is building a QaaS service with ethics and accessibility, with the aim of democratizing access to quantum services.

In this Q&A, Dynex co-founder Daniela Herrmann discusses the company’s approach to QaaS, the importance of an ethical framework and GPU-based emulation versus quantum hardware.

Enter Quantum: What is unique about Dynex’s QaaS offering?

Daniela Herrmann: We started the project in 2020 and brought to market a subscription model for enterprises to access quantum compute. We were aware that it's the next mega-cycle of computing but not everybody’s ready for it, they need a lot of handholding and education.

We didn’t set out to just provide a subscription then things would just happen, we built the entire infrastructure behind it, with a decentralized network of 150,000 GPUs feeding into the computation that the clients then want to do.

Why is ethical use of Dynex’s QaaS important and how do you enforce it?

Related:How Businesses Can Safeguard Operations From Quantum Threat

We created a layer-one blockchain that serves the ethical imprint that we want. Anyone who wants to use that compute must declare their ethical principles and we can help them with that. That declaration is then immutably recorded in the blockchain.

That's new to the industry, but from everything we’ve seen from cutting-edge technology, any founder should be responsible, because we know what these technologies are capable of and have to be careful. We also incorporate flagging mechanisms that flag when codes are suspicious. If the flagging mechanism is triggered, it will immediately go to the Dynex moonshots ethical committee for review. If it results in rejection, we have another loop to make sure we didn't miss anything.

What is Dynex’s quantum compute architecture?

We use GPU-based emulation. It’s not comparable with a monumental IBM quantum computer, that's a different strategy of wanting to push quantum computing through.

I think the evolution of the quantum industry will fork into actual quantum computers and emulations because for many applications you just do not need quantum machines. But for others, some military applications, for example, you're probably better off with a quantum machine.

Related:Why Early Adoption of Quantum Is Key for Financial Institutions

What industries is Dynex targeting?

It’s a very exciting moment because we're going from the horse to the car. Quantum will not just be for highly intellectual research activities, this is going to make life easier for a lot of young, small, mid-cap companies that are, for example, exploring the optimization of supply chains combined with weather prediction.

In the bigger industries, you have wind tunnel calculations for the automotive industry and RNA folding in medicine. At such sweet spots of changing a paradigm, you always also see a change in players. I cannot say who will be the players of the future, but I do think that there will be different players in the next mega-cycle of computing than we all think.

How is Dynex supporting the next generation of quantum users?

We’ve just launched a textbook, which we built into a program at Sona College of Technology in India, where we have 23 top-notch students going through a 10-week program. Then we would love to expand the access to 6 million people because it should become mainstream.

Quantum is not something for an elite, this is for the mainstream. If a little business wants to grow internationally and can suddenly afford to think in a new way, it has completely different opportunities.

I’m from Switzerland, so I'm always a bit strict with the word democratization. Logically, it's not full democracy, because if it was, we would just give it out as open source. That is not the case. Making it mainstream is probably a better word for what we're doing.

How is QaaS supporting the evolution of AI?

The massive handling behind ChatGPT or the video rendering for Netflix, for example, are huge computational challenges, and they're also extremely cost-intensive. Quantum will logically play an important role because by having AI go hand-in-hand with quantum, you get much shorter processing cycles and better precision.

There are a lot of initiatives now around organic intelligence, where you create an AI from scratch that is not initially connected to the internet, to create a good grounding in humanity's goals before letting it out of the house. Because they're currently a bit like a trash bin.

How can you see quantum and QaaS evolving in the next five years?

There’s something of a dialog happening about whether quantum is coming in 15 years or whether it’s already here, so there could be a surprise coming.

With the attempts we're making and what we’ve heard from the market, larger companies will start implementing their quantum strategies soon and they will need reputable integration partners that can help them through the process of quantum readiness.

What are Dynex’s plans to incorporate quantum hardware?

Our innovation journey is to get a room-temperature silicon quantum chip out this year and embed it into the network we created. If we sold the chip we’d scrap our ethics, so it will be embedded as a node in our decentralized computing model. But right now, we don’t want to go for a machine realization.