Posts

The Power Of Artificial Intelligence In The Mortgage Industry

These days, it’s hard to miss the buzz about artificial intelligence (AI) and its impact on industries such as health care, automotive, education, financial services and retail, to name just a few. From the ability to diagnose diseases – to the development of driverless cars – the potential applications of AI are extraordinary. In our daily lives, we already are experiencing the use of AI when we communicate with customer-service chat bots, ask Apple’s Siri for information, perform Google searches, or use navigation apps to help avoid traffic, as a few examples.

Despite all the recent discourse about AI, this technology is certainly not new. There are countless examples of AI use over the past several decades, including the reliance of commercial jet flights on AI to power autopilot, and internet bots that index web pages. But the more recent interest, innovation and investment in AI are due to a combination of factors – including greatly increased computational power, big data, greater infrastructure speed and scale, open source technologies and advancements in machine learning techniques.

Featured Sponsors:

 

 
And today, the mortgage industry is able to reap the benefits of this incredible technology. For example, HeavyWater, which was recently acquired by Black Knight, is a provider of AI and machine learning-based capabilities specific to the financial services industry. The company has already built a platform that completes business tasks using synthetic read-and-comprehend analysis and conclusion skills, and applied these capabilities to the loan origination process.

What is Machine Learning?

The terms “machine learning” and “artificial intelligence” are often used interchangeably, however, there is a distinction between the two. Using a very broad definition, artificial intelligence replicates human reasoning through learning, problem-solving and pattern recognition. Machine learning is a subset of AI and is a process by which AI deepens its knowledge through continually performing tasks and processing information.

Let’s consider a simple, industry-specific example. AI-powered machine learning enables technology to “remember” standardized forms. For example, it can review thousands of paystubs and determine exactly where the pertinent income data is located. When the system comes across a paystub that presents an anomaly, it will apply its previously gained understanding to infer the location of the income data needed. Once the technology receives feedback that its inference was correct, it incorporates that information into its knowledge base. The next time it comes across that type of paystub, the system will automatically know where to find the pertinent data.

Featured Sponsors:

 
Machine learning also leverages big data to gain insights. The more data that is collected and reviewed, the better machine learning solutions become at making predictions.

Applying AI and Machine Learning to Reduce Costs and Turn Times

AI and machine learning already can make a difference in two of the biggest challenges faced today by mortgage originators: costs and cycle times. With the ability to read, comprehend, and draw conclusions based on context, AI and machine learning can perform operational functions more efficiently and at scale.

In fact, machine learning can work on many of the labor-intensive, “stare and compare” tasks performed by humans – such as verifying income, assets and insurance coverage. Machine learning is used to perform these manual activities much faster and more accurately than humans – a task that takes employees hours to complete can be reduced to just seconds with machine learning.

By automating manual routines, machine learning not only expedites the origination process, but also increases volume. While humans can only work a certain number of hours before mistakes begin occurring, machine learning has no limits to the time or energy it can spend performing these tasks. By increasing loan processing volume and reducing mistakes, imagine how machine learning can drive down origination costs – and risk.

AI-powered systems enable processors and underwriters to dedicate more time to addressing exceptions and solving problems, which will improve transaction turn times. Also, AI can help avoid last-minute delays by prompting a lender’s staff to take early action when there is an issue, keeping the origination process moving forward. Additionally, by delegating work to AI-powered technology, a lender’s staff can focus on delivering a more positive and personalized consumer experience.

Featured Sponsors:

 
As AI and machine learning are used to perform manual, repetitive tasks, allowing mortgage professionals to work on more value-added responsibilities, lenders can increase their focus on their company’s growth strategies. As they scale and reduce the cost per loan by keeping staffing levels flat, lenders can invest more in product development, marketing, infrastructure, and other growth-oriented initiatives.

Additional Applications of AI

AI can also leverage visual recognition to image and index a wide variety of documents that are typically reviewed by processors and underwriters, such as tax returns, W-2s, property titles and appraisals. A lender could even use AI and machine learning to better manage vendors. Based on past performance and cost, AI could provide recommendations on which vendors would be optimal for each loan going through the origination process.

Voice-integrated AI brings further opportunities to create efficiencies. This technology could look at information under review, evaluate results and automatically employ interactive communication bots to advise employees of any issue that may need attention. Additionally, via a conversational interface, processors and underwriters could ask for information they need – just as we use virtual assistants like Apple Siri, Amazon Alexa or Microsoft Cortana to get answers. These capabilities certainly could help move a loan through the origination process faster.

Leveraging AI to Enhance Customer Service

Of course, most of us have experienced first-hand how AI is applied in retail to deliver a more personalized consumer experience. For example, when we shop online, we receive targeted product recommendations the next time we visit that site; or receive faster service though chat bots.

To help personalize and enhance the borrower’s experience, lenders can leverage voice capabilities. A mortgage virtual assistant that engages customers by answering questions, walking them through the application process and even offering advice could be employed using voice-integrated AI.

Impact on Jobs

When the subject of AI in the workplace is discussed, it inevitably raises questions about its impact on jobs. Will jobs be lost as a result of these technological advancements?

There is no perfect answer to this question since the utilization of AI is different from company to company. But, it seems certain that future skill sets will be required to support this shifting technological paradigm. As it applies to the mortgage industry today, however, AI can enable professionals to spend less time on remedial work, becoming knowledge workers instead of task executors, and provide additional value to a company.

The Future Is Limitless

AI and machine learning offer tremendous potential to advance the mortgage industry, and we are just beginning to experience the technology’s capabilities. As AI-powered systems ingest more data and perform an increasing number of tasks though machine learning and other techniques, the possibilities are unlimited.

Imagine the power of AI as it learns to handle the entire point-of-sale process and speaks to an applicant directly through a mobile phone; or as it systematically searches a lender’s portfolio for qualified prospects and offers a customized home equity loan or line of credit, and so on. As we all know, the average cost to originate a mortgage loan is exceptionally high – today it is nearly $8,500 according to the Mortgage Bankers Association’s Quarterly Mortgage Bankers Performance Report, and the typical time to close a loan is 41 days. Any opportunities to reduce costs and increase process efficiencies will add value to lenders and consumes.

What’s more, the transformative power of AI doesn’t stop in the originations space. Servicers will also be able to reap the benefits of this advanced technology. For example, the technology could learn how to detect risk and any compliance issues before they occur, enhance loss mitigation decisioning, provide voice integration capabilities to help staff work faster and smarter, and so on. What’s amazing is that these examples only scratch the surface.

Of course, human interaction will always be needed to originate and service loans, as people will still decide how they want to leverage technology and determine the problems that must be solved. Humans must also still play an active role in loan decisioning, identifying which kind of data to consider and determining risk appetite. Furthermore, research indicates that despite all the advances in point-of-sale technology, consumers still want the comfort of human interaction at some point in the process of purchasing what is most likely their largest and most important investment.

AI and machine learning offer great promise and will likely usher in a new era of production excellence. Lenders that take advantage of this advanced technology will be choosing a bold new way to address origination costs, improve turn times and transform their origination processes to support a brighter, more successful future.

About The Author

Soofi Safavi

Previously CEO of HeavyWater Inc., the mortgage-focused Artificial Intelligence (AI) provider recently acquired by Black Knight, Inc, Soofi Safavi now serves as Managing Director of Black Knight’s Applied AI group, bringing leading-edge AI and computing capabilities to the Black Knight product portfolio. With over 20 years of experience in mortgage and banking technology, and deep expertise in IT strategy, architecture and machine learning, Soofi is uniquely suited to discuss AI’s role in the mortgage industry.

Not All Banking Interactions Of The Future Will Be Handled By Bots

With Mastercard Inc. set to release its first-quarter financial results, and Finovate Spring 2018 approaching next week, fintech continues to lead the disruption charge in banking. US banks remain under pressure to stay ahead of what’s in store during the next wave of innovation, which is being led by AI advancements in the industry.

Featured Sponsors:

 

 
Gartner’s research shows the global banking industry will spend $519 billion on IT in 2018, up 4.1 percent year over year from $499 billion in 2017. Features like chatbots are changing banks’ relationships with customers, and digital players are increasingly integrating social media to interact with clients.

But despite the increase in technology adoption in the sector, the need for human contact persists, with voice recognition and chatbot technology still in its infancy.

Featured Sponsors:

 
Traditional banks must heavily invest in upgrading contact center environments to accommodate anticipated customer needs, in conjunction with embracing advanced technology platforms.

Banks can reap the benefits of AI and automation to improve interactions between staff and customers, says Intelenet Global Services, whose tech innovations for banks have seen them win recognition in the 2017 IDC Fintech Rankings.

Featured Sponsors:

 
Tony Antenucci, VP of Banking, Financial Services and Insurance for Intelenet Global Services, comments: “The way customers interact with their bank in the U.S. is changing, as branches close down and more people switch to online banking, communicating through contact centers via phone or online if they need more help.

“Keeping up with new technology is vital to keep the processes behind these operations running smoothly and maintaining a relationship between banks and their customers. AI tools can help personalize banking for the customer, and automated processes free up staff to focus more on customer service and complex question resolution.

Tony continues: “For example, voice-recognition programs can automatically recognize an individual by their voice, predicting the subject of a call before the customer has to explain it. Each customer can be automatically sent to the right department, or the person they spoke to before, without having to be passed between contact staff.

“This dramatically improves the experience for the consumer. One leading bank was able to reduce the average handling time for customer calls by 40 percent.

“Banks will need to keep up this lead in innovative technology, to fend off growing competition for customers from Fintech challengers.”

About The Author

Tony Garritano

Tony Garritano is chairman and founder at PROGRESS in Lending Association. As a speaker Tony has worked hard to inform executives about how technology should be a tool used to further business objectives. For over 10 years he has worked as a journalist, researcher and speaker in the mortgage technology space. Starting this association was the next step for someone like Tony, who has dedicated his career to providing mortgage executives with the information needed to make informed technology decisions. He can be reached via e-mail at tony@progressinlending.com.

The Risks And Rewards Of Artificial Intelligence For Lenders

In looking at this, the recent debut of self-driving cars could transform a stressful commute into an opportunity to tackle emails and reading lists while making suburban long-distance travel great again. Americans are poised to gain more than 100 hours per year in free time by relinquishing the wheel to smart cars. The downside, though, lies in inevitable vulnerabilities like security threats, job loss, and environmental impact. Is the reward of AI worth the risk?

Artificial Intelligence is broadly defined as a computer’s ability to perform tasks normally requiring human intelligence. Ever since Alan Turing developed a test to determine a machine’s ability to exhibit intelligent behavior in 1950, roboticists and scientists have sought to pass it. In 2014, a computer program called “Eugene Goostman” succeeded by convincing 33% of the human judges that it, too, was human. Since then, companies like SAP, General Electric, and MasterCard have utilized machine learning and artificial intelligence (MLAI) to identify trends and insights, make predictions, and influence business decisions.

Artificial intelligence offers new opportunities to revolutionize operations in the financial services industry. Machine learning can process terabytes of data in seconds – volume which a horde of humans with older machines or methods could not process in a lifetime. The sheer power of modern computing in assessing an increasing number of variables with speed and accuracy gives a financial institution the ability to have rapid and strategic insights about customer behavior, reporting errors, and risk patterns. In the area of loan decisioning, this should lead to faster loan origination, fewer compliance problems with regulatory fines, and more inclusive lending overall.

Featured Sponsors:

 

We wanted to know: What are the risks and rewards of AI in lending, and is it an inevitable next step for compliance management?

Bob Birmingham, CCO of ZestFinance, and Dr. Anurag Agarwal, President of RiskExec at Asurity Technologies, explore.

Which Discrimination Would You Prefer: Human Or Machine?

Anurag Agarwal: With machine learning, the decision-making is supposedly agnostic to overt biases. Humans, by definition, are free thinking but with biases that interfere with decision-making, especially in lending scenarios.

Bob Birmingham: AI is not a product but a solution, another approach to problem solving using advanced mathematics, analytics, and data. It’s not a magic fix but can remove some of the human discretion that leads to discrimination.

Featured Sponsors:

AA: The problem is that computers could reinforce societal disadvantages by relying too heavily on data patterns. When a machine imports data, it creates an algorithm but cannot explain its methodology. It may introduce variations in the mix without understanding the broader implications.

Say you are trying to determine the risk of repayment for a borrower. The machine identifies physical appearance as a decision point. Based on data patterns, it detects that blue eyed individuals are more prone to timely repayment of loans. Noticing a correlation, it elevates that decision point to a higher influence in future lending. By taking an apparently agnostic data element and making a correlation it has in essence created discrimination because it doesn’t understand that blue eyes are mostly representative of Caucasians and the societal implications of making that borrower preference as a result.

BB: We’ve been here before as an industry. When regression modeling first came out it was confusing, flawed, and its accuracy questioned. At the end of the day, it was actually a more accurate process than purely judgmental underwriting because it followed a clear set of instructions to find answers. MLAI creates an opportunity for continued improvement to mitigate and eliminate discriminatory lending practices.

Risk: Machines can learn the wrong lessons from data.

Reward: Machines don’t have overt biases.

Alternative Data

AA: AI is very new. We don’t yet know how to regulate it or what its long-term impact is. As soon as data began generating, we followed this “go forth and multiply” pattern. Now we have more data than we know what to do with. Also, a lot of this data is unstructured.

Featured Sponsors:

BB: Alternative modeling and alternative data are often lumped together, but that shouldn’t always be the case. MLAI alternative modeling may be used on the same data that current regression models run on and provide lift. Alternative data may be used with your existing models too but is more often used in conjunction with alternative modeling due to MLAI’s ability to handle large data sets. MLAI also has the ability to identify missing, erroneous, or wrong data and solve problems related to the inaccuracies.

AA: Alternative data, such as Facebook profiles, Twitter feeds, and LinkedIn pages, are already used in employment. HR departments use this publicly available information to determine if you are a qualified candidate. By knowing this in advance, you can change your online behavior to skew these data points in your favor. That’s the big question with alternative data in the lending space, along with privacy and transparency issues. We don’t know who gets to manipulate the data, where it comes from, or who has access to it. There’s no established system or standardized data points, which means most companies must follow their own proprietary lending rules, making it a regulatory free-for-all.

BB: While there are caveats, the rewards significantly outweigh the risks. There are many underserved individuals, businesses, and micro-borrowers with little or no credit that could benefit from alternative data. If an applicant doesn’t have good credit, or any credit at all, lenders can use MLAI techniques to paint a richer portrait about the borrower’s reliability using nontraditional factors like e-commerce histories, phone bills, and purchasing records. MLAI can open up the credit market, measure patterns, and fill in the data gaps, giving lenders a more holistic view of an applicant. Alternative data doesn’t have to be “creepy” data and doesn’t have to be social media data. Responsible and transparent use of alternative data to expand access should be encouraged.

Risk: The use of alternative data in credit underwriting raises privacy, transparency, and data integrity issues.

Reward: Alternative data can increase financial inclusion by granting access to capital to individuals and businesses with no traditional credit history.

Hackers vs. Trackers

AA: As we saw with Equifax, any time data is automated through the cloud there is the risk of data hacking. Now we are asking: who is responsible for protecting all of this data? Using such a detailed lending profile increases the necessity for data privacy and security.

BB: These are risks but the industry has always been responsible for protecting sensitive data.  The good news is, the more information we have about a borrower, the quicker we can identify errors and anomalous behavior. This is a great consumer benefit. Using the same Equifax example, imagine if we could say “this person was affected, and these actions are very different from their previous activity. This is a red flag that their information was stolen.”

Banks raise red flags when uncharacteristically large payments are made or a card is used in a different country, but imagine how much more effective these alert systems could be with additional insights into an individual’s unique behavior patterns.

Risk: More data in the cloud means a higher risk to consumers in the event of a data hack.

Reward: Additional data can identify unusual behavior quicker, which is a benefit to consumers.

The Regulator’s Dilemma

AA: Regulators have a big job ahead of them figuring out how to regulate companies using this information. Many of these companies aren’t sure how to use it themselves.

BB: Initially, users of MLAI in high stakes applications have an obligation to educate the public, create a set of best practices for their industries, and be transparent. Financial inclusion is something regulators support and this technology can help lower the barriers to entry.

AA: In late 2017,the CFPB issued a no-action letter to Upstart Network, an online lending platform that uses both traditional and alternative data to evaluate consumer loan applications. The terms of the letter required Upstart to share data with the CFPB about its decision-making processes, consumer risk mitigation, and methods for expanding access to credit for traditionally underserved populations.

By studying these companies, regulators can better understand the impact on credit in general, on traditionally underserved populations, and on the application of compliance management systems.

Risk: Regulators are still learning about AI and how to properly monitor it.

Rewards: Regulators support consumers and want to make access to credit more inclusive.

The elephant in the room: Jobs.

BB: Typically, Financial Investigative Units (FIUs) are looking at alerts 24/7, researching a person, tracking where money is going, and determining if they should file a suspicious activity report. It’s the banks biggest compliance cost and their highest area of employee attrition with numbers ranging from 15 – 35% turnover in the FIUs. With AI, FIUs can focus on stopping financial crimes rather than toggling back and forth between 15 screens and parcelling through tons of information. This should free compliance personnel up to do higher value, more rewarding work in addition to driving more efficient outcomes for their organizations.

AA: I prefer a human loan officer over an automated machine-learning system. If something happened six months ago that caused your credit to go down, you can explain that to a loan officer. How can you convey that to an automated system? Those intangibles make human interaction necessary. I believe there is something valuable in interpersonal interactions that can never be captured in a truly automated fashion.

Risk: Headcounts in certain departments that are reliant on manual processes could decrease.

Reward: New and more rewarding job responsibilities will result in less turnover, but ultimately there is no replacement for interpersonal interaction for certain positions.

Closing Arguments

BB: Currently, the industry operates with a “look back” approach for Fair Lending, which doesn’t really work. The whole process of defending and explaining discrimination after a model is put into production feels outdated. Today, we can and should build AI models to proactively remove discrimination before ever putting a model into production.

AA: This technology is still very new. Medium to big size lenders should let the technology emerge and wait to see what the regulatory landscape looks like. Regulators still have a ways to go before any of us will fully understand what the future looks like for AI in the lending space.

Augmented Intelligence

One of the first introductions of artificial intelligence to the general population came in 2011 when Watson competed on Jeopardy. Ken Jennings and Brad Rutter were arguably the best players the show had produced over its decades-long lifetime. In total, they had walked away with more than $5 million in prize winnings, a testament not only to the breadth and depth of their knowledge, but their strategic savvy with category selection and wagering. Watson, a computer system developed by IBM, was capable of answering questions posed in natural language. Watson had access to 200 million pages of structured and unstructured content, but was not connected to the Internet. Watson consistently outperformed its human opponents on the game’s signaling device, but had trouble in a few categories, notably those having short clues containing only a few words. The key here was the development of a natural language processor that would become the foundation for numerous future applications like Siri. But while its rapid responses to questions may have struck many as robotic, Watson was not a robot in the traditional sense. Robots are machine built to carry out physical actions and may or may not be designed to approximate the human form. I am sure many of you remember the TV series, ‘The Jetsons’ with Rosie, the humanoid robot maid and housekeeper. Or maybe not, because that series was on in the early 1960s.

Featured Sponsors:

 

 
‘In Search of a Robot More Like Us’ was a 2011 New Your Times science article written by John Markoff. He stated that:

The robotics pioneer Rodney Brooks often begins speeches by reaching into his pocket, fiddling with some loose change, finding a quarter, pulling it out and twirling it in his fingers. The task requires hardly any thought.” But as Dr. Brooks points out, training a robot to do it is a vastly harder problem for artificial intelligence researchers than IBM’s celebrated victory on Jeopardy…. Although robots have made great strides in manufacturing, where tasks are repetitive, they are still no match for humans, who can grasp things and move about effortlessly in the physical world. Designing a robot to mimic the basic capabilities of motion and perception would be revolutionary, researchers say, with applications stretching from care for the elderly to returning overseas manufacturing operations to the United States.

Featured Sponsors:

 
So, let’s leave the discussion about robots for another time. Instead, I’ll focus on defining augmented intelligence and differentiating it from artificial intelligence. It’s more than a question of semantics. Artificial intelligence, perhaps from its popular culture use in general and its science fiction use in particular, can conjure up images of the sentient machines with personal agendas. It suggests a culture where, at least in some part, humans are no longer required to make decisions. Some industry experts believe that the term artificial intelligence can create more negative speculation about the future than hope.

Whatis.com defines augmented intelligence as an alternative conceptualization of artificial intelligence that focuses on AI’s assistive role, emphasizing the fact that it is designed to enhance human intelligence rather than replace it. An alternative label for artificial intelligence also reflects the current state of technology and research more accurately.

Featured Sponsors:

 
According to an article by Athar Afzal, “We’ve transitioned from an agricultural-dominated society to the industrial revolution – and now to a more data-driven economy. What we’ve witnessed during each of these stages is some form of mechanics or machinery developed to augment our performance, thereby improving our outcome…. The world has a lot of opportunity to gain and make our lives better with augmented intelligence – it’ll make our lives far smoother and more enjoyable. I invite everyone to view Ginni Rometty’s speech at the World Economic Forum.”

Researchers and marketers hope the term augmented intelligence, which has a more neutral connotation, will help people understand that AI will simply improve products and services, not supplant the people who use them.

While a sophisticated AI program is certainly capable of making a decision after analyzing patterns in large data sets, that decision is only as good as the data that human beings gave the programming to use. The choice of the word augmented, which means “to improve,” reinforces the role human intelligence plays when using machine learning and deep learning algorithms to discover relationships and solve problems. I’ve summarized some definitions by Jean-Albert Eude below.

Machine learning is a type of artificial intelligence (AI) that allows software applications to become more accurate in predicting outcomes without being explicitly programmed. The basic premise of machine learning is to build algorithms that can receive input data and use statistical analysis to predict an output value within an acceptable range. The processes involved in machine learning are similar to that of data mining and predictive modeling. Both require searching through data to look for patterns and adjusting program actions accordingly.

Deep learning is an aspect of artificial intelligence (AI) that is concerned with emulating the learning approach that human beings use to gain certain types of knowledge. At its simplest, deep learning can be thought of as a way to automate predictive analytics. While traditional machine learning algorithms are linear, deep learning algorithms are stacked in a hierarchy of increasing complexity and abstraction. Each algorithm in the hierarchy applies a non-linear transformation on its input and uses what it learns to create a statistical model as output. Iterations continue until the output has reached an acceptable level of accuracy. The number of processing layers through which data must pass is what inspired the label “deep.” The advantage of deep learning is that the program builds the feature set by itself without supervision. This is not only faster, it is usually more accurate. In order to achieve an acceptable level of accuracy, deep learning programs require access to immense amounts of training data and processing power, neither of which were easily available to programmers until the era of big data and cloud computing.

The value of such augmented predictive analytics to a segment of the economy as dependent on data as the mortgage industry is obvious. What is also obvious, unfortunately, is that we may be among the last to seat ourselves at the technology table.

Often, an early title or tag line for a concept or theory evolves over time as others develop their ideas and work toward a solution. In the mortgage industry, the concept of paperless mortgages was proposed in the early 1990s to reduce and/or eliminate what some conceived as unnecessary paper and to improve the overall experience for the consumer. Along the way we started referring to it as an electronic mortgage (e-mortgage) and now it is the digital mortgage, an all-inclusive data and documents packaged in a format for both human and machine consumption. That will certainly achieve the initial objective to eliminate paper and improve the consumer experience. The operational benefits extend from origination all the way through to the secondary market.

But going digital without building the internal architectures to capitalize on data-driven support technology is like going to a 3D movie, but not putting on the 3D glasses to watch it. If we don’t keep moving our own finish line, we risk being trampled by those with a longer view of the race.

About The Author

Roger Gudobba

Roger Gudobba is passionate about the importance of quality data and its role in improving the mortgage process. He is an industry thought leader and chief executive officer at PROGRESS in Lending Association. Roger has over 30 years of mortgage experience and an active participant in the Mortgage Industry Standards Maintenance Organization (MISMO) for 17 years. He was a Mortgage Banking Technology All-Star in 2005. He was the recipient of Mortgage Technology Magazine’s Steve Fraser Visionary Award in 2004 and the Lasting Impact Award in 2008. Roger can be reached at rgudobba@compliancesystems.com.

The Good, The Bad And The Reality Of AI

website-pdf-download

When I was getting my Masters in Business Administration, one of my professors lectured about the future of business. One example he used was from a futurist’s ideas on the factory of the future. According to this individual the factory of the future would have two employees: a man and a dog. The man would be there to feed the dog and the dog would be there to make sure the man did not touch anything.

While this tale may seem laughingly far-fetched, conversations held at the recent technology conference seemed to indicate that this scenario is within the realm of possibility. Numerous discussions were held about the use of artificial intelligence (“AI”) within the mortgage banking industry and ranged from rules-based programs to utilizing programs such as optical character recognition (“OCR”) and validating document collection to produce the first “bots” within the industry. But is this really AI or is real artificial intelligence the actual use of computers that can emulate human thinking processes and contain human drives such as hunger, power and self-preservation?

Controversy abounds on the subject and not just with potential users but among the most advanced thinkers in this area. Their thoughts and beliefs are wide-ranging from the concept that AI will likely play a part in human extinction to those that believe it will improve people’s lives and give them more family and leisure time. An article in April’s Vanity Fair magazine quotes Elon Musk, the developer of Tesla cars and cost-efficient rockets allowing for the settlement of Mars, as believing AI is humanity’s “biggest existential threat.” On the opposite end of the scale, Ray Kurzweil, a futurist, has predicted that we are only 28 years away from the point where AI will far exceed human intelligence and humans will merge with this super intelligent program to create the hybrid beings of the future.

Featured Sponsors:

 

 
While arguments continue at this theoretical level, most of these individuals agree that one of the greatest drawbacks to the full use of artificial intelligence is the interaction currently necessary between humans and these tools. In other words, for these programs to work, a human must verbally ask a question, make a statement or key in information. This problem goes away however with the merger of biological intelligence (human thought) and machine intelligence. To accomplish this, companies are currently working on an injectable mesh, called a neural lace, into the brain that can flash data from your brain wordlessly to your digital device or to the cloud, thereby creating unlimited computing power.

With the on-going merger of AI into businesses there is also concern about how it will be managed and controlled. Current public policy on AI is largely undetermined and the software is largely unregulated. Some of the biggest technology companies have taken it upon themselves to develop a partnership on the subject in order to explore the full range of issues, including ethical concerns. The European Union is also deeply concerned and is considering such legal issues as whether robots have personhood or should be considered more like slaves as found in Roman law.

But the question overriding all of these issues appears to be what exactly is artificial intelligence? Is it simply a bot-like program that runs rules that do simple labor intensive work or is it actually the ability of machines to think as humans and take over the entire workload of any business. And if so, what does this actually mean to employees in those businesses? Will their jobs be replaced with robots that not only collect information and compile data for a rules-based engine but actually make decisions based on the data fed into the machine’s intelligence.

Featured Sponsors:

 
Most importantly to mortgage lenders, what does all this mean to our industry? While the fun and intellectual stimulation that comes from brain-storming these ideas generates lots of enthusiasm, if these concepts become reality, we need to be prepared to utilize them to our advantage and not be thwarted by extensive costs and back room operations.

One way to envision how these AI programs will impact the industry is to look at what has been happening in similar operations. An article by Penny Crossman in the March 16, 2017 American Banker entitled “All the Ways AI will slash Wall Street jobs” gives some insights into what we might expect. According to the author, Opimas, a capital markets consulting firm, projects that by the year 2025 artificial intelligent technologies will reduce employees in the capital markets profession by 230,000 people. Furthermore, spending for AI-related technologies is expected to be more $1.5 billion and will reach $2.8 billion annually by 2021, just four years from now. This number does not even include the start-ups that capital market firms will invest in during this period. All of this expense is expected to be offset by a 28% improvement in their cost-to-income ratios.

So where are these programs being placed? The first functions being replaced by AI technology are process-oriented jobs. These jobs are actually being replaced by lower level AI functions that are programmed to do such things as look up documents, find data and compare multiple data sets.   In addition to these process oriented jobs, those whose function is to conduct analytics on the data are also being replaced with such technology as machine learning functions. In this “deep learning” technology, AI programs digest large volumes of real-time data within a very short period of time and then “learn” to find patterns that provide insight and direction at a speed humans can’t begin to match.

Another area of capital markets feeling the impact of AI is front office sales personnel. Since initiating AI technology in this area there has been a 20% to 30% drop in headcount. In addition, many jobs in the middle and back offices are also feeling the impact. Since the majority of these jobs are processes that are connected by human manual intervention AI that brings with it image recognition can replace this human activity.

Featured Sponsors:

 
Compliance concerns that resulted in significant headcount increases are now being taken over by AI programs that validate specific documentation and provide a more holistic view of the regulatory risk and organization compliance trends. This is one area where IBM’s Watson is proving extremely valuable.

The implementation of AI in capital markets gives an excellent overview of how this technology can be implemented in the mortgage industry. Currently, we use some lower level rules-based programs to conduct underwriting as well as OCR usage in some back-office functions. Applying the applications discussed previously, many, if not most of the job functions being conducted today by humans could in fact be replaced with future AI programs.

One good example is the use of AI to replace loan officers for taking applications and collecting data. The Rocket Mortgage program in use today by Quicken Loans is just one example. Furthermore, most of the data collection and organization process that is labeled “processing” is also easily replaced with existing sites that offer independent validation of the information utilized in making decisions.

An area that is also ripe for AI application is the title and closing function. Using OCR, data comparison and document production, can easily be completed while the risk of mistakes or problems at the closing table could be handled in real time.

Back room operations can also be easily incorporated into AI functionality since it is very much a data recognition, document collection and validation effort. Post-closing functions which now take time and massive amounts of human labor can not only be streamlined, but the data collected can be used to revise and improve the processes themselves.

While what may appear to be already included as an AI function, underwriting is actually where some of the best deep learning artificial intelligence is applicable. Since 1995 we have been using rule-based technology to conduct what we considered AI, but instead are simply automated underwriting programs. Deep learning AI offers the industry the solution that has plagued it since its inception, that of identifying the true performance risk of loans.

Today’s credit risk function continues to use static attributes to develop, expand and or shrink credit policy without knowing the potential impact on any individual applicant. This credit risk stalemate has resulted in lending programs that reject applicants that may in fact prove to be credit-worthy borrowers. This can easily be seen in such programs as affordable housing and minority lending. Using deep learning, rather than simply applying standard credit policy to an application, AI can conduct an analysis of the applicant in comparison to all probabilities of performance and decide to approve or reject. In other words, credit evaluations would be individualized for every applicant. In addition, performance probability would be the yardstick by which pricing is tabulated.

Servicing is of course, primarily manual back-office functions that AI can address. Once again a deep learning application can contain any information on taxes, insurances and related issues, transfer funds if and when needed as well as provide an escrow analysis, tax statements and individual billing statements.

Just this brief recap easily demonstrates the value of bringing AI into the mortgage lending environment. The question is “at what cost?”   There is no doubt that the advancements in AI would have the same negative impact on mortgage lending employees. In fact, the annual convention might just turn out to be a dog and a man. However, as shown above, mortgage lending is not the only industry that will feel the same impact.

There are of course risks. One significant enough to delay implementation of AI by some firms is the risk that the technological intelligence could misinterpret input information and make decisions based on that information that would be disastrous to the company. One such example has already been experienced by Wall Street to a small degree when a mention of Anne Hathaway in the news resulted in a bump in Berkshire-Hathaway’s stock. Now known as the “Hathaway Effect”, companies are implementing practices such as running validation scenarios and are placing restrictions and stops on critical process points. This of course requires human oversight and runs headlong into the issue of AI and the reduction of human jobs in the industry.

The discussion over job elimination and creation however needs to be a much broader discussion around the impact of AI in the economy overall as well as in our industry. We have to think about what the massive reductions in employment opportunities means and what type of jobs will be created as current competencies are becoming less relevant and those trained in AI technology become harder to find. This change from the use of human intelligence to artificial intelligence is similar to the change undergone by those individuals who today are labeled “white working class” as large scale manufacturing replaced humans with more advanced technology and those individuals who were educated to do manual and factory related jobs became unemployable. Those that were smart enough to take advanced training and education in the field have found new jobs, but those that haven’t are left feeling angry and disenfranchised.

New skill sets that will be in demand for AI revolve around software engineering and data science. This is of course a given. But there will also be the need for a hybrid of business and digital skills which involves individuals who are knowledgeable about the business, who understand the digital environment and know how to benefit the business by continually improving its digital footprint.

While AI continues to advance, and becomes more accepted in the industry, it would be wise of us to think of when, where and how it becomes the most advantageous to the human side of the business equation. Using AI also means finding what our function as humans involves. Are we the masters of the technology or will we become evolve to a position seen by Steve Woznick when he said, “I now feed my dog steak because I see humans being the pets for robots and I want my robots to believe pets eat steak.”

About The Author

Rebecca Walzak

rjbWalzak Consulting, Inc. was founded and is led by Rebecca Walzak, a leader in operational risk management programs in all areas of the consumer lending industry. In addition to consulting experience in mortgage banking, student lending and other types of consumer lending, she has hands on practical experience in these organizations as well as having held numerous positions from top to bottom of the consumer lending industry over the past 25 years.

The AI Era Is Here – Pt. 2

website-pdf-download

In the latest issue for Fortune, Erin Griffith examines the investment trends in AI (Artificial Intelligence) technology and poses the question: is AI an overhyped fad or a revolution? She writes, “There’s an easy way to tell when the hype around a technology trend has peaked. 1) Are the smartest venture capitalists complaining about valuations? 2) Are big tech companies snapping up start-ups so young they can barely be considered real businesses? 3) Are Fortune 500 executives talking about their [insert trend here] strategy? If the answer to any of these questions is yes, congratulations! You’ve identified a fad.”

Featured Sponsors:

 

Of course, most revolutions look like fads in their early days—because they are. Distinguishing between those fads that will fade and those that will become the norm in the long term can be difficult, and as in all things, hindsight has a much higher success rate than foresight when it comes to identifying the winners and the losers. So what data might guide such an evaluation?

The research firm CB Insights recently reported that in 2016 there were 658 venture capital deals in the AI sector. In 2016, that amounted to $5 billion in startup funding deals, a significant increase from $589 million in 2012. International Data Corporation projects worldwide revenue from artificial intelligence and cognitive systems to be $47 billion in 2020, up from $8 billion in 2016.

Featured Sponsors:

CB Insights selected 100 of the most promising artificial intelligence startups globally from a pool of 1,650 candidates based on factors like financing history, investor quality, and momentum. A look at the top 50 shows that AI is surging worldwide with 20% located outside the United States. They cover a wide range of market segments: core AI, FinTech, auto, health care, commerce, CRM, cyber-security, robotics, business intelligence, and text analysis and generation.

Interestingly, the fact that AI does not necessarily intersect with established business cases has not proven to be a hurdle to investment. “These are not businesses,” says John Somorjai, executive vice president of corporate development at Salesforce, which has acquired a handful of AI companies. “These [deals] are about technology and talent.”

Featured Sponsors:

The development of artificial intelligence has inspired both fascination and dread.

In 1955, the term AI represented the concept of autonomous systems modeled on the structure of the human brain. At the same time, other researchers were tackling a different problem: finding patterns in what was then considered great volumes of data and making proper selections, or decisions, based on that data. In 1956 William Ross Ashby wrote in his Introduction to Cybernetics that “…what is commonly referred to as ‘intellectual power’ may be equivalent to ‘power of appropriate selection’.” This was not intended to as “artificial intelligence” in the way we typically understand it, and in fact was labeled as the inverse: IA, or Intelligence Augmentation. If this model sounds suspiciously familiar, it is because today’s AI systems are constructed on the IA paradigm. Our real-world applications, including language processing, machine learning, and human-computer interaction are based on IA—data pattern recognition and appropriate decision making—and as such, they augment our capacity to understand what is happening in the complex world around us. While the term “AI” became the label of choice for such technology, it is an ironic misnomer.

Let’s look at the “Why You Should Let Artificial Intelligence Creep Into Your Business” article in the March, 2017 issue of Inc magazine for some definitions:

How AI works: problem solving: Unlike traditional computing, which delivers precise solutions within defined parameters, AI, sometimes referred to as cognitive computing, teaches itself how to solve problems. “Instead of delivering specificity, AI-centric programming generates millions of solutions, evaluating each for efficacy and then choosing the most viable and optimal ones,” says Amir Husain, CEO and founder of SparkCognition.

What it does better: data diving: Manually finding your target customer, by searching and poring through income-level, interest-based, and geographical data, is labor-intensive and time-consuming. AI cuts to the chase. “For example, using a feed of three key pieces of information that the entrepreneur provides; a brief product description text, images and a price range; an AI system can zip through social media and other online outlets, looking for correlations between product and digital conversations,” says Husain. If you give it the green light, AI’s natural language processing technology then writes and sends a sales pitch, notes transmission times, and analyzes feedback. “You can almost hear an AI system going, Aha! I’ve cracked the code.” says Husain, adding that AI constantly optimizes itself by making slight changes to the message.

Where it works: practical apps: One key reason for AI’s upsurge is entrepreneurs’ free or inexpensive access to libraries such as IBM Watson, Goggle TensorFlow, and Microsoft Azure. These application programming interfaces (APIs) allow coders to build AI apps without starting from scratch. Husain expects to see a proliferation of AI-centric marketing, sales and other service startups focused on small and medium-size businesses.

Let’s look at some specific examples from the same article.

Call Centers: The biggest misconception about AI is that it’s robots with human faces sitting at remote desks. “AI is nothing more than an add-on technology, spice and flair, to an otherwise conventional system, such as a traditional travel-reservation site that, because of AI can now converse with a human,” says Bruce W. Porter, an AI researcher and computer science professor at the University of Texas, Austin. Porter emphasizes that future breakthroughs will not be 100 percent AI. “AI will likely provide a 10 percent product or service performance boost,” he says. That is, in fact, huge. Firms that fail to make the leap, he says, may fail to have customers.

Information Retrieval: Not all searches are as simple as typing a few keywords and having Google take over. Entrepreneurs often need more in-depth and complicated excavations for patent and trademark data, for example and that, in turn, involves an often-hefty legal budget to pay a highly-trained human to do. Porter foresees within five years many companies offering services to consumers who have no experience in AI or specific knowledge fields. They’ll be able to conduct their own AI based data retrieval. Count on industry disruption, he says, as this type of AI application will leapfrog current data-retrieval-service providers.

Contract Generation: Because it’s able to generate natural language, AI is an exceptional tool for helping entrepreneurs assemble contracts, as opposed to buying them off the shelf at, say LegalZoom. AI applications will converse with – by text and, ultimately, voice – and tease information out of humans that will become components of formal agreements, such as details about fee payments and product returns. Porter anticipates users will pay to access cloud-based AI computer systems to produce such documents. AI-centric startups, because they don’t require a human in the loop and won’t need to hire staffers, can offer their services at a very low cost, especially given an anticipated large volume of customers and business competition.

AI can displace humans, but it can’t replace them.

Leaders of every industry and institution are sprinting to become digital. Who will win? The answer is clear: It will be the companies and the products that make the best use of data. And the ones that make the best use of data will likely be the ones that use AI to gain efficiencies in data analysis and decision making.

About The Author

Roger Gudobba

Roger Gudobba is passionate about the importance of quality data and its role in improving the mortgage process. He is an industry thought leader and chief executive officer at PROGRESS in Lending Association. Roger has over 30 years of mortgage experience and an active participant in the Mortgage Industry Standards Maintenance Organization (MISMO) for 17 years. He was a Mortgage Banking Technology All-Star in 2005. He was the recipient of Mortgage Technology Magazine’s Steve Fraser Visionary Award in 2004 and the Lasting Impact Award in 2008. Roger can be reached at rgudobba@compliancesystems.com.

A Look Into The Future

I recently conducted an industry roundtable. It was made up of a good group of executives. They were very opinionated and thoughtful. I learned a lot on that call.

But one thing stuck out at me. I asked about the future of mortgage lending technology and Jeff Bradford, Founder and CEO of Bradford Technologies, Inc. said this:

“It’s all about deep learning. Deep learning is like artificial intelligence,” Bradford explained. “There is so much work going on in this space right now and it’s open source technology, so anyone can use it. There’s a guy that used the software to create a self-driving car inside of a month. Big companies have been trying to develop a self-driving car for years. Once the software learns something it, doesn’t make a mistake.”

Featured Sponsors:

 

Jeff Bradford is a renowned expert in appraisal analytics and accuracy in collateral valuations, and the mastermind behind computer-aided appraising, the most far-reaching and significant advance in the appraisal segment in decades. He frequently presents and speaks at industry events, on topics that include technology, valuation processes and valuation standards. He is a strong proponent of open industry standards and was one of the chief architects of the MISMO Appraisal XML standard.

Jeff started his career teaching early developers how to program the Macintosh at Apple Computer. He also specialized in computer-aided engineering at Structural Dynamics Research, and performed and managed computer aided analysis projects for FMC Central Engineering Labs.

Eventually he started Bradford Technologies. Bradford Technologies is an innovator of valuation tools and solutions for residential appraisers. The company pioneered computer-aided appraising, was the first to incorporate statistical support in both mainstream and alternative valuation products, and currently provides one of the most adopted technologies for residential appraisers. AppraisalWorld, the company’s online appraiser community with over 20,000 members provides services focused on building trust and reliability in the appraisal industry.

Jeff is a very smart guy. I’ve come to know and admire his intellect over the years, so when he talked about deep learning I had to know more. I came across this article that talked about when Ray Kurzweil met Google CEO Larry Page. As the story goes, he wasn’t looking for a job. A respected inventor who’s become a machine-intelligence futurist, Kurzweil wanted to discuss his upcoming book How to Create a Mind. He told Page, who had read an early draft, that he wanted to start a company to develop his ideas about how to build a truly intelligent computer: one that could understand language and then make inferences and decisions on its own.

It quickly became obvious that such an effort would require nothing less than Google-scale data and computing power. “I could try to give you some access to it,” Page told Kurzweil. “But it’s going to be very difficult to do that for an independent company.” So Page suggested that Kurzweil, who had never held a job anywhere but his own companies, join Google instead. It didn’t take Kurzweil long to make up his mind: in January he started working for Google as a director of engineering. “This is the culmination of literally 50 years of my focus on artificial intelligence,” he says.

Kurzweil was attracted not just by Google’s computing resources but also by the startling progress the company has made in a branch of AI called deep learning. Deep-learning software attempts to mimic the activity in layers of neurons in the neocortex, the wrinkly 80 percent of the brain where thinking occurs. The software learns, in a very real sense, to recognize patterns in digital representations of sounds, images, and other data.

The basic idea—that software can simulate the neocortex’s large array of neurons in an artificial “neural network”—is decades old, and it has led to as many disappointments as breakthroughs. But because of improvements in mathematical formulas and increasingly powerful computers, computer scientists can now model many more layers of virtual neurons than ever before.

With this greater depth, they are producing remarkable advances in speech and image recognition. Last June, a Google deep-learning system that had been shown 10 million images from YouTube videos proved almost twice as good as any previous image recognition effort at identifying objects such as cats. Google also used the technology to cut the error rate on speech recognition in its latest Android mobile software. In October, Microsoft chief research officer Rick Rashid wowed attendees at a lecture in China with a demonstration of speech software that transcribed his spoken words into English text with an error rate of 7 percent, translated them into Chinese-language text, and then simulated his own voice uttering them in Mandarin. That same month, a team of three graduate students and two professors won a contest held by Merck to identify molecules that could lead to new drugs. The group used deep learning to zero in on the molecules most likely to bind to their targets.

Isn’t this just amazing? Just imagine the mortgage applications for this technology. For years we’ve talked about using technology to automate mortgage processes or forms, but this technology could actually revolutionize all of mortgage lending and how we think about processing a loan. Hopefully someone in our space will catch on to deep learning and use it to transform mortgage lending for the better.

About The Author

Tony Garritano

Tony Garritano is chairman and founder at PROGRESS in Lending Association. As a speaker Tony has worked hard to inform executives about how technology should be a tool used to further business objectives. For over 10 years he has worked as a journalist, researcher and speaker in the mortgage technology space. Starting this association was the next step for someone like Tony, who has dedicated his career to providing mortgage executives with the information needed to make informed technology decisions. He can be reached via e-mail at tony@progressinlending.com.