“We’re probably in the second or third inning.”
That is Andrew Los Status report on the progress of artificial intelligence (AI), big data and machine learning in finance.
1. Prejudices
Lo said that applying machine learning in areas akin to consumer credit risk management is actually just a primary step, but now the industry is attempting to use machine learning tools to higher understand human behavior.
The big query is whether or not machine learning will ultimately reinforce all of our existing human biases. Agrawal actually doesn’t imagine that.
“If we had had this conversation a few years ago, the question of bias wouldn’t even have come up,” he said. “Everyone was worried about training their models. Now that we’ve achieved value in a number of applications, we’ve started to worry about things like bias.”
So where does the priority about bias come from?
“We train our models on different types of human data,” Agrawal explained. “So if the human data has a bias, not only does the AI ​​learn that bias, but it can potentially amplify the bias if it thinks that will increase its ability to optimize or make more effective predictions.”
But AI can be used to reduce bias. Agrawal cited a study from the University of Chicago study Researchers developed AI programs that not only mimicked the bail decisions of human judges, but in addition more accurately predicted the chance of flight.
2. Economy and wealth distribution
There is little doubt that AI increases productivity. But will AI create a jobs crisis by making human staff obsolete? Agrawal believes that individuals are nervous because we do not know where the brand new jobs will come from, nor will we know if those that lose their jobs later of their careers will give you the option to reskill for these recent positions.
Innovation is occurring so quickly today that we do not know whether reskilling programs might be as effective as they were up to now – even for younger staff who’ve the time and opportunities to really participate.
The other problem is wealth distribution. Will the introduction of artificial intelligence result in greater wealth concentration?
“I would say that almost all economists agree that this will definitely lead to economic growth and therefore an overall increase in welfare for society,” Agrawal said. “But there is a split among economists on what this means for distribution. Some of us are very concerned about distribution.”
3. Regulations
According to Lo, there are various opportunities for brand spanking new varieties of data within the financial sector.
“We still need to know so rather more in regards to the financial ecosystem, especially how [inputs] interact with one another over time in a stochastic environment,” he said. “Machine learning can use large amounts of knowledge to discover relationships that we were previously unaware of. So I believe we’ll see much faster progress in all of those AI methods which have thus far been applied to a much smaller data set.”
Agrawal raised a related concern: “In regulated industries like finance, healthcare and transportation, the obstacle for many of them is not data. We cannot deploy it because of regulatory hurdles.”
Lo agreed that regulations could potentially hinder progress.
“There are a number of complex problems that we don’t really know how to regulate right now,” he said. “A good example is autonomous vehicles. Currently, the laws are designed so that if someone causes an accident and kills another passenger or pedestrian, they are responsible. But if an AI is responsible for a death, who is responsible? Until we solve that aspect of regulation, we won’t be able to make the progress that we could.”
AI and machine learning for everybody
How can finance professionals develop their skills in machine learning, big data and artificial intelligence?
“There are a lot of really useful courses you can take to get up to speed in these areas,” Lo said. “But it just takes a certain amount of time, effort and interest.”
The younger generation is best positioned on this regard, said Lo. In fact, today’s youth have more trust in human-machine relationships, said Agrawal, because they’ll simply spend more time with computers, mobile devices, etc.
As Lo explained on the outset, we’re still on the very starting of the applying of those recent technologies in finance. There are high hopes that they’ll increase productivity and result in higher profits, but at the identical time there are concerns in regards to the potential impact on wealth concentration and employment.
Nevertheless, concerns that the introduction of artificial intelligence and large data could reinforce human biases could also be exaggerated, and the potential obstacles posed by regulations could also be underestimated.
However, given the inevitable adoption of AI in finance and beyond, finance professionals cannot afford to be ignorant.
If you liked this post, don’t forget to subscribe.
