No credit card required
Browse credit cards from a variety of issuers to see if there's a better card for you.
Artificial intelligence is changing credit cards and banking
"In general, the more data AI-powered software can gather, the more effective it will be. Credit card issuers, with a world of consumer data at their fingertips, stand in a strong position to provide their customers with the best options for their lifestyles."
I work in the non-financial data analytics field and almost every single time we run a machine learning simulation with new updated data from new sources we are amazed at what turns out to be a good predictor of something. Usually it's "Wow, never would have guessed that one." from many in the room. There's just far too many interdependencies for a human to comprehend.
That's why everyone wants everything they can get a hold of from consumers today. Daily Fitbit data, GPS data from phones (apps), what time of day someone watches TV (cable box), when they turn up the thermostat (Nest), when they eat (smart fridge)....really getting insane, and it's only going to get more invasive.
Yeah, it's going to be an awkward implementation in the financial markets; as it gets more efficient, we might also basically get red-lining all over again. There was a somewhat recent study that I was reading about in last week's Economist, and when they ran the data through, most of those efficiencies went to white borrowers.
I don't know what the right solution is, but at some level we have to be able to feed the algorithms good data, and I'm not entirely sure the data sets aren't stacked against some segments of the population.
@Revelate wrote:Yeah, it's going to be an awkward implementation in the financial markets; as it gets more efficient, we might also basically get red-lining all over again.
Am I correct in assuming that the ultimate 'efficiency' as it applies to credit lending would be finding people most likely to carry a balance month to month and always pay it on time? Or do they make more money from high-spend transactors?
I can think of so many things AI/ML could track right now that might end up as a good predictor of 'integrity'. Social media posts, obviously, but maybe in the future things like data from the CAN bus in a car that shows a driving pattern more often than not associated with people who don't pay their bills on time. (Swerve too much and your FICO score goes down a couple points? We probably shouldn't laugh.)
I’m also interested in this from a hobbyist perspective.
Here is my take. The AI never operates in vacuum. They operate under the specter of well....greed. Let me explain. If the data shows some groups should get credit and other groups should not that’s fine on its surface. But what usually happens is humans can’t leave group B alone. Strictly left up to the AI that group would be severely limited in its ability to get credit.
But humans see that group as an opportunity for profit. High APR and high fees. The attitude is “well they should be happy with whatever they get”. But the AI would look at the high APR, high fees, no grace period and be able to predict a high failure rate. Which the humans don’t care about because they will make enough in predatory fees to offset the charge offs.
If we do this right we will eventually be able to point crowd sources AI’s at products offered by institutions. “Credit card XYZ from bank ABC has a 97% chance of causing you to file bankruptcy in 3 years.....avoid”.
@Anonymous wrote:Am I correct in assuming that the ultimate 'efficiency' as it applies to credit lending would be finding people most likely to carry a balance month to month and always pay it on time? Or do they make more money from high-spend transactors?
While this behavior is likely the best for maintiang good credit, and less likely a risk. Hence better scores, and probably income.
There's also the fact that mnay Banks make a lot of Money on those people who consistantly pay late. I'm sure Amex loves their Swipe fees, but I'm sure they like the late fees too.
As for the AI invasion into my perosnal matters, I don't need my CC deciding what's right for me to buy. This is why I'd decided not to get a Smart Fridge, Nest etc. While all cool and convenient items that could make life easier, I just don't need/want the invasion/tracking.
It's all fun and games until the AI turns on us all.
@Anonymous wrote:But humans see that group as an opportunity for profit. High APR and high fees. The attitude is “well they should be happy with whatever they get”.
This is why I left software development in the financial sector to work with analyzing data from machines.
That sector sees everyone like this: Black Mirror - Nosedive episode excerpt
"The episode is set in a world where people can rate each other from one to five stars for every interaction they have, which can impact their socioeconomic status. Lacie is a young woman overly obsessed with her ratings; she finds an opportunity to elevate her ratings greatly and move into a more luxurious residence after being chosen by her popular childhood friend as the maid of honour for her wedding. Her obsession leads to several mishaps on her journey to the wedding that culminate in a rapid reduction in her ratings. "