Publisher and reporter with Silicon Hills News
Social media has caused an explosion in the amount of data generated by people and businesses in the last five years.
Yet, data without analysis is worthless.
That has led to the rise of artificial intelligence and machine learning coupled with the input of humans to make sense of streams of data, according to experts on Argo Digital’s panel Sunday at South by Southwest on “How Data and Machine Learning/AI Affect Risk Transfer in the 21st Century.”
Jason Abbruzzese, business reporter with Mashable, moderated the panel for Argo Digital, an emerging insure tech practice within property and casualty carrier Argo Group, based in San Antonio.
By analyzing the data, all businesses are trying to make themselves smarter, said Andy Breen, senior vice president for Argo Digital and adjunct professor at NYU Stern School of Business and one of the panelists.
Until recently, the tools did not exist to sift through all the data and extract insights that make a business operate better, he said. Every day, people and companies generate all kinds of streams of new data on Google, Facebook, Twitter, Instagram and more.
“There was definitely a time when it was very uncomfortable, but now we have the tools to deal with it,” he said.
The fact that we have all this data out there is not very interesting, said Andrew Bocskocsky, co-founder and CEO of Grata Data and another panelist. Companies have been using data as long as data has been available, he said.
“When we start having buy in from industry it starts getting interesting,” he said. “We can generate value and get results everyone is happy with.”
Sambit Sahu, adjunct professor of computer science at Columbia University, and another panelist, said to get great analysis, companies need great data. He is also a data/telco analytics research scientist who was involved with the IBM Watson program.
“Garbage in, garbage out,” Sahu said. “The data collection is only as good as the data quality.”
That’s why businesses like Argo are looking for new ways to collect data.
For example, with a restaurant, Argo is learning new ways to assess risk, Breen said. It is now deploying sensors, drones and other internet of things devices to create new data streams continually, he said.
“Getting data constantly through the stream changes the conversation,” he said. “We can then talk to them about risk management.”
And data collection works hand in hand with humans, Breen said. A lot of data is unstructured and requires humans to figure it out, he said.
“It is the combination of algorithms helping people be better assessors of risks or people be better decision makers,” Breen said. “Let’s let algorithms do what algorithms are good at which is complex math and working with really large data sets. Things like that. And have people do the things they are good at like sales and negotiations and other types of things like that. When we reach the singularity, if we ever reach the singularity, we’re not going to have machines that are going to be at that level.”
Six month ago, Breen looked at the underwriting process, collecting data from ten different data sources to assess risk, and built a tool that can collect all that data and digest that, he said. That focuses Argo’s employees on what they are good at which is risk assessment, he said.
The cost has come down on data analysis computing power which has led to more data analysis and insights, Sahu said. That has accelerated the artificial intelligence part of the analysis, he said.
The data still cannot eliminate risk in the insurance industry, Breen said.
“No matter how much better the data is and no matter how much we underwrite there is still going to be hurricanes, tornados and floods and other things to deal with,” he said. “I would even be an advocate of leveraging the playing field for the data. Everyone should have equal access to equal amounts and open data. That would actually be better generally for the industry as a whole. They can decide how much risk to take on.”
The next five years is going to be a period of “choppiness” and transition with a trend toward artificial intelligence as a platform that creates its own neural networks to solve problems, Breen said.
“Maybe five, maybe ten years, someone can go in and say here’s my problem and the neural network will actually build itself,” he said. “That’s starting today. We’re not there yet, but it’s happening.”
Another trend is insurance companies are partnering with startups to leverage their technology and get into this space, Bocskocsky said.
“You’re giving data and expertise and they are giving you their knowledge and technology,” he said.
When it comes to privacy, all the data is out there, Breen said.
“Personally, I disengaged with Facebook three or four years ago,” he said. “I felt like it was wrong.”
Content on Facebook is developed to target behaviors and emotions, he said.
“You’re giving up tons of valuable data every time you’re on Facebook and you like something,” he said.
Editor’s note: Argo is an advertiser with Silicon Hills News