AI will decide whether or not you can afford your home
Your service provider has their own AI. We individuals need our own AI representation in the programmatic exchange.
AI will decide whether or not you can afford your home. Algorithms are becoming more and more fundamental in all decisions in the world. How do we ensure that the direction we are building is the one we want to build? With all the issues around data bias and transparency, can we trust algorithms with our key decisions? Yet, without any true power on our side, how can we negotiate with an at times omnipotent AI decision maker?
I was privileged to participate in the Data Series roundtable call “Data and ML in Lending” organized by our friends at Open Ocean VC. With Niklas Rosenberg moderating, and participants Abhijit Akerkar from TBC Bank Group, Roberts Bernans of Nordigen, Mark M from Humn.AI we discussed the current trends surrounding new data and technology in loan underwriting.
Black boxes are on the way out, intelligent AI needs to be explainable and societies in the future include not only humans but also algorithms. That needs new solutions and new perspectives.
Transparency and Explainability Is Key
A key trend on its own, black boxes offer no sunlight for decision making; and transparency is crucial in order to trace back key decisions (e.g., whether or not a couple got the mortgage) to the factors that influenced the decision. While other dimensions such as what data the models are trained on, what goals they have, are important, explainability on its own is a core trait.
Right before the pandemic, Prifina organized an event in San Francisco “Towards New Customer Relationships: Empowering Individuals With Data” where we had Pranay Podge from EazyML talk about the importance of explainability of AI, where explainability is itself not a mystery but plain english that we can understand. How else are we going to understand it, let alone trust it?
Starting from Individuals Themselves
Discussions on lending nearly always start with the lender. However, it is also interesting to look at it from the perspective of the borrower, the individual person or persons. With the Prifina community, we’ve been exploring examples of reverse matching, where the individual themselves can set up their own ask (in this case, a mortgage application) and then the suitable lenders can be connected to the individual for further discussions and processing.
From the individual’s point of view, if they were to approach a number of lenders directly, they would get stuck in the lenders’ processes and silos, without necessarily knowing that lender is the most suitable overall for them. Marketplaces and auction mechanisms are powerful as has been demonstrated time and time again, yet in financial services too often they start with only the service provider in mind.
AI Is Also an Ecosystem Participant
On the Liberty. Equality. Data. podcast, we’ve also discussed how algorithms have become a fully functioning participant of society and how their importance is growing. In a recent episode, we talked about how to regulate algorithms in society and how to find the balance between safeguards and fostering innovation and competitiveness. We’ve also discussed the importance of not only having algorithms, but having micromanagers for those algorithms (thank you, Peter).
Individuals themselves have an important role in the future, but so do data and algorithms, and the various other technological advances. What’s possibly fundamental, is that these technologies find their way onto the individual’s side. This can be as simple as your own AI listing (transparently) and suggesting what mortgage lenders seem most appropriate to your needs, without sharing any data. AI can build user empowerment, if it truly represents the person.
There is a large difference, whether or not a service provider is utilizing your personal data or if it is actually you, using it for your direct benefit. Algorithms will surely decide whether or not you can afford the house, but whose algorithm it is matters. If the bank has their algorithm, where is yours?
We need software on the individual’s side to foster more opportunities and open the market, and in the case of mortgages, find the algorithms worth talking to in the first place.
Connect With Us and Stay in Touch
Prifina is building resources for developers to help create new apps that run on top of user-held data. No back-end needed. Individual users can connect their data sources to their personal data cloud and get everyday value from their data.
Follow us on Twitter, Medium, LinkedIn, and Facebook or listen to our podcast. Join our Facebook group Liberty. Equality. Data. where we share notes about Prifina’s progress. You can also explore our Github channel and join us at Slack.
