Looking back at 2022, it was obvious that Artificial Intelligence made incredible strides. These breakthroughs came in the form of NLP and Computer Vision to Generative AI and Explainable AI. We caught up with AI/ML experts from JP Morgan & Chase, UBS, University of Greenwich, Cornell University, and Fidelity Investments to find out about the most popular trends in Artificial Intelligence and Machine Learning in finance in 2023. Here's what they had to say:
Adam McMurchie - Lead Data, DevOps & Cloud Engineer, Contractor
"Personally, I think the trending use of unexplainable AI is both interesting and deeply concerning, we are currently undergoing the Oil Rush era of AI, where we are driven by the desire to obtain short-term value without considering the long-term implications.
Unexplainably isn't simply a biproduct of complex design, but rather an inherent flaw in conventional neural networks which produce black-box models and whilst hundreds of tech companies have worked to address this, the fundamental issue still persists.
This isn’t hypothetical, significant damage has already been incurred in the finance space from poorly aligned AIs, and whilst it’s a drop in the bucket compared to the benefits reaped – it won’t seem that way to those on the receiving end.
Just like electric cars in the early 1900s, alternatives to unexplainable AI exist, they may even have a greater potential than the best deep learning algorithm, but market conditions and the academic landscape are failing to incentivize innovation in this area.
“Why build a better hammer that no one knows how to use, when this one works just fine?” One could argue, that the current hammer will knock the house down eventually."
Luke Vilain - Director of AI Ethics, UBS
"Fairness within algorithmic or data systems - without care we will continue to propagate historical biases, and unless we understand how the bias of designers influences data products we'll not be as incentivized to ensure teams are diverse.
Our society is becoming more and more data driven, meaning there is ever more impact from unfair bias on us as individuals or as a society"
Georgios Samakovitis - Professor of Fintech, University of Greenwich
"My focus is on developments in GraphML and federated learning (FL), particularly because of the powerful affordances the two technologies provide in areas where data sensitivity is core, as is the case in AML and FinCrime, my primary research concern.
Lately, a more acute demand for leveraging transaction and other custodial data for knowledge discovery brings GraphML and FL into stronger focus for FinTechs. Their combined capabilities for privacy-preserving knowledge discovery and sharing, promise, perhaps for the first time, cost-aware, scalable AML solutions that are also commensurate with regulatory requirements for data privacy – in turn, an area of increased attention in FinTech and beyond."
Darian Nwankwo - Ph.D. Candidate, Cornell University
"The trends that have captured my interest are in language models. Outside of my research in AI, I consider myself an amateur philosopher; in particular, I spend some time thinking about the philosophy of mind and how language has a tendency to be the primary mechanism behind how we elaborate our thoughts.
Language models are becoming increasingly relevant since most people have the capacity to communicate using language; the simplest of queries to these models would prompt any individual to ask the question “how?” Whether you’re knowledgeable of the inner workings of artificial intelligence or not, you’ll be interacting with it—much like electricity."
Upal Sen - VP - Squad Lead/Product Owner AI, Fidelity Investments
"Large language models and conversational AI. The interest that ChatGPT has been able to generate from the AI and business community is a testament to the power of conversational AI. There is a lot more to do here and work on improving accuracy and effectiveness of such models. The upside for our industry though is tremendous with the possibility of completely revolutionizing human-based customer service and financial advice."
These leading experts will be joining us at the AI in Finance Summit New York on April 20-21, 2023, where they will be discussing the challenges of AI in Finance in more detail and how to overcome them.
Standard Rate ticket sale for AI in Finance Summit New York ends on Friday, April 7, so secure your place today to save $200.