Wells Fargo CIO on the Early Days of Quantum Computing

In this interview from the AI Summit New York, Chuck Martin talks to Peter Tsahalis, Technology Sr. Executive & CIO at Wells Fargo. Peter outlines the emerging but essential use cases of quantum computing.  

This forms part of the AI in Finance series of content you can find here: AI in Finance

Introduction:  

Chuck Martin: 
Hi, this is Chuck Martin at the AI Summit in New York City. I have with me here from Wells Fargo, Peter Chisales. So you're dealing with quantum. Where is quantum in the grand scheme? People keep saying it's coming. Is it here yet? 

Peter Tsahalis: 
It's just been born, let's put it that way. We are just now starting to see hardware, and the qubit race begin amongst different vendors. But the utility of it is the real question. What can you really do with 100 qubits, 200 qubits, 300 qubits? 

Challenges and Hardware Development in Quantum Computing 

Peter Tsahalis: 
How much noise or error comes out of the implementations that you attempt to do? So a lot of large vendors in this space, IBM being one of the biggest one that's recently expanded their roadmap, predicting qubit counts and error correction rates. 

So that's what we keep a close eye on. But in terms of, at least in financial services, practical applications, we can't quite implement them yet, but we can experiment and at least prove that they will be valuable someday. 

Integration of Quantum and Classical Computing 

Chuck Martin: 
So do you see it as integrated with current computing or as a separate thing? 

Peter Tsahalis: 
The way most of us envision it is, quantum computing will be another form of compute power. Just like today, you can get classical compute power, you can get GPUs for doing AI work. Quantum workloads will be a type of workload that if you can develop your algorithms appropriately, you can send certain type of work to quantum machines for processing, but there will be, I think for the foreseeable future. 

Quantum Cryptography and Cybersecurity Concerns 

Chuck Martin: 
Berenice Baker, the editor of Enter Quantum, one of our publications, she's always seeing cyber security as a really big deal. What's the play there? 

Peter Tsahalis: 
The area of post-quantum cryptography was the first thing that everyone started worrying about when it came to quantum tech. Theoretically, when you have a quantum computer with the right capability, number of qubits, noise level, you could in real-time snap encryption, you know, that would take weeks, months of a supercomputer today. 

Post-Quantum Cryptography Preparations 

Peter Tsahalis: 
So how do you prepare for that, right? And how much lead time do you need to prepare for that? And what do you do? So a lot of companies have started making investments in the area of PQC. 

The Real Risk: Long-Term Data Exposure 

Chuck Martin: 
So how does a company deal with the bad guys who are capturing this encrypted data waiting for their quantum computing to decrypt it? 

Peter Tsahalis: 
The same way we do already... It's not about protecting data today from an attack today, it's about protecting data that may have been stolen today from an attack that might happen ten years from now on that same data. 

The Timeline for Quantum's Practical Use in Financial Services 

Chuck Martin: 
So, where do you look at the timeline? The CTO of FedEx who spoke here yesterday, he said that yes, they see quantum coming, but they think it's more of a five-year thing. Some people say, well, we're doing something now. 

Peter Tsahalis: 
It's hard to predict these things. Five years from now, if you look at some of the vendor roadmaps, the size and scale of quantum hardware and the expected error correction and mitigation techniques, I think should allow some early use cases to form. 

Quantum's Potential in Scientific Research and Optimization Problems 

Peter Tsahalis: 
Scientific applications are already entering a time of quantum utility where certain scientific research where you're modelling a particular element... But when we're talking about optimization problems, predictive research, predictive problem space, you need to map huge data volume into the system, huge circuit depth. So we're not quite there yet. 

Industries Beyond Financial Services That Can Benefit from Quantum 

Chuck Martin: 
What do you see as some of the most important industries that should be in this? I mean, financial service is obvious. Are there other industries that you think should be or are involved in Quantum? 

Peter Tsahalis: 
If you're in an industry where it would be helpful to you to predict something based on past data, or to profile something like a customer, or to profile something like fraud information, then yes, there's a use case. 
A lot of the power of quantum isn't necessarily just quantum in isolation. It's the marriage of quantum and the intersection of quantum with AI, blockchain. 

Quantum Machine Learning and Future AI Algorithms 

Peter Tsahalis: 
A lot of the research we do at Wells Fargo is how do we take AI algorithmic work that's applied classically and translate that into a quantum space. So quantum machine learning, not just regular machine learning. Quantum recurrent neural networks, not just regular one. 

Outlook for Quantum in the Next Year 

Chuck Martin: 
So a year from now, we're sitting here having this conversation around quantum. What will we be talking about in a year? 

Peter Tsahalis: 
A year from now we'll have a lot more knowledge on things that we think will work and not work. A year from now we'll have more partners in the space, more companies making inroads, but in terms of practically applied quantum, I don't think a year from now we'll see much difference. 

Find Out More

Don't miss out on the latest insight and innovations that will be shared by the experts at the 2024 New York Summit - get your pass now!

Or take a look at some of the highlights from the 2023 event and see what you can be part of.