Banks see artificial intelligence in their future, but are slow to invest in it

Most financial firms believe artificial intelligence will have a huge impact on their own business and the overall industry in the next few years. If that’s so, investment and integration need to begin fairly soon. And the reality is, most banks are far from doing so.

Below are five charts that show the current state of AI in U.S. financial institutions; what banks’ AI plans are; and why they’ve been slow to move on them, despite the hype.

No plans
In a Celent survey of banks, the only technologies being used are fraud analysis and risk detection and natural language processing, by 14 percent and 5 percent of respondents, respectively. For all the hype around chatbots, none of the banks have fully deployed one, although 9 percent are running pilots. Celent doesn’t identify which banks participated in the survey but it’s safe to assume those piloting chatbots are among the largest, like Bank of America’s erica as well as Capital One and Chase, who have also reported bot pilots.

Some banks are running pilots with RPA and Natural Language Generation, but most are still just considering different technologies or haven’t made any plans.

Few are investing in AI
Just 9 percent of banks, presumably the same ones that are running pilots and seeing positive results, plan to invest more than 50 percent in AI in the next year than they did in the last one. Another 23 percent of banks indicated they planned to invest more than 25 percent more. Thirty-two percent indicated their investment levels would remain the same, but another 32 percent indicated they’re not investing at all.

“A lot felt it was expensive, a lot of people have a general lack of understanding,” said Stephen Greer, an analyst at Celent and co-author of the AI report. “There’s such a perception and understanding that it will have an impact, but there’s a limited number of institutions doing a lot with it.”

Five percent of banks said they planned to invest less than 25 percent in AI in the coming year than they did in the last.

Expense and security
Fifty-eight percent of banks indicated that AI might be too expensive to implement and deploy. The same percentage listed security as a concern — although 13 percent indicated they weren’t concerned about security at all.

Most striking for Greer was the high percentage (31 percent) of respondents who said ethical considerations weren’t at all a concern. Eliminating bias, abusing AI and displacing people’s jobs all fall under the ethics category — and they’re some of the largest sources of uncertainty for most consumers who don’t work with the technology hands on.

Where does it go?
According to PwC, most firms in banking, capital markets and wealth management say they’ll rely more on human judgment than machine algorithms to inform their biggest decisions.

Further, according to Celent, banks aren’t very optimistic about AI’s applications for their customers. They seem most excited about using it for fraud detection — particularly as fraudulent activity gets more sophisticated.

“I don’t get the sense most of the industry was acutely aware of how [AI] will happen or why it will evolve,” Greer said.

For some AI technologies, it takes time to digest the customer data they need and to have humans teach them how to react appropriately. Over time, the hope is that AI can reduce banks’ cost of serving customers and improve customer interactions. It’s hard to see that in these early stages, when AI is still in the learning stage. However, banks showed they don’t expect AI to handle more than 30 percent of their front or back end operations.

 

‘We’re not there yet’: USAA’s Darrius Jones on security concerns in the next big channel — voice

People could soon be doing their banking over voice-activated channels. But there are major issues around security and privacy to iron out first.

On Wednesday, USAA began piloting an Alexa skill for Amazon home assistant devices that lets customers check balances, review spending history and get other account insights based on their transactions. USAA is keen on letting Alexa read back customers’ financial data, but it’s not ready to let Alexa make payments, said Darrius Jones, assistant vp at USAA Labs, a division of USAA.

Many industries, not just financial services, are getting concerned about Amazon inserting itself between them and their customers. Banks and fintech startups are interested in using voice platforms to reach customers, but data and identity security and privacy concerns loom.

Tearsheet caught up with Jones about the pilot, its relationship with Amazon and staying ahead of customers’ security needs. Answers have been edited for length and clarity.

What are some of the security challenges of this pilot?
Understanding Amazon’s role in security versus our role. Privacy is another. When you have one of these devices and you plug it in, it has to listen. That’s part of the challenge and what makes them work. It’s what you’re allowed to do with the things you hear that people are now kind of going back and forth on.

How do you mean?
You don’t want your information spewed out into the ether when anyone can be in your house and ask a question.

Or move money around.
We have not put any money movement capabilities on the platform at this point. [It] is not something our skill will accommodate because we’re not comfortable with the state of security for money movement on the platform. How do you do this seamlessly and securely? We’re just not there yet.

Is Amazon a competitor or a partner?
In this conversation we’re definitely in a partnership. We’ve had to asked them to help us better understand the technology platform, we’ve had to help them better understand our regulatory requirements.

Does Amazon keep USAA customers’ data?
Amazon only has access to what the member provides during the interaction with Alexa while using the USAA skill. We use OAuth 2.0 to provide the member with the ability to see what permissions Amazon will be granted and give them the power to decide whether to grant that permission, which they can also revoke at any time. Amazon knows the question that the member asks Alexa and the response that is provided, but not the raw data used to formulate the response. All the transaction data is USAA-owned data.

Customers often care less about privacy than they think and more about speed and convenience.
We’ve enabled secure key, a six-digit key enabled with the Alexa skill that has needs to be uttered upon invoking the USAA skill. Only once you do that will you be able to get personalized spending information and balances. It was something Amazon asked for, but even the way we implement it — having it directly on the Amazon platform, where you have to set it up to determine whether to keep it on or off — is another useful usage pattern we focus on.

Inside the development of Erica, Bank of America’s AI-powered bot

Brands talk a lot about artificial intelligence and its potential to change the way they interact with customers, but few really understand what exactly that potential is or what it takes to make it real.

However, when Bank of America got a sense of the vision for its AI-enabled “digital assistant,” called erica, it didn’t take the bank long to gather the resources necessary to make her real.

“We realized that to do what we wanted, there would have to be a huge investment of time, energy and resources to make it happen,” said Henry Agusti, digital banking executive. the bank’s digital banking executive. “We committed to doing that very early on — about 9 or 10 months ago.”

The bot is now in the beta testing phase.

Since last fall, Bank of America has pooled together a team of more than 100 people dedicated to building erica, a voice- and chat-driven product designed to help customers manage their financial lives using predictive analytics and cognitive messaging. The “vast majority” of the team are people who were working at B of A before it started the idea of erica, Agusti said, since a lot of the functions designed into the digital assistant will be deeply integrated with a lot of the bank’s existing back-end systems.

Erica is really like a personal banker. Only she takes the relationship to a whole new level of personal; she’ll have all of the customer’s data along with thousands of other machine learning inputs that will give her even more information about the customer. In the near term, customers will be able to ask questions, execute transactions and look at balances.

As erica learns from those transactions, she’ll be able to offer more to customers, like insights and advice; and from there, she can start to give advice more proactively instead of merely reciting the information she has.

“In the next 18 to 24 months, the path is about how much functionality we can give erica,” Agusti said. “We really want to make sure she’s able to help clients and customers execute on a lot of functionalities from day to day. That requires a deep level of [technology] integration.”

It’s a far cry from the rest of the industry: While 85 percent of financial institutions surveyed by Celent agree that AI will have a significant impact on banking in the next three to five years, 32 percent indicated they’re making any direct investment into AI technology. Half of the banks surveyed indicated that the expense of integrating AI is a major concern.

Bank of America declined to give details of its financial investment in AI or in erica.

“It looks like an aggressive timeline, but it we still need to see the finished product. Advanced AI or natural language processing skills take time to learn and mature before they are ready to go live,” said Celent analyst Stephen Greer. “Banks using even the most advanced AI like Watson need time to train it.”

Erica will be available to the bank’s associates in a few months, and in November she’ll offer the ability to provide real-time insights and advice, like flagging a dip in a customer’s FICO score or suggesting payment plans based on changes in subscriptions.

“We teach erica how to deal with one customer request but we’re finding customers are asking for other things we never even imagined,” Agusti said.

The work is as much on the front end as the back end, Agusti said. It’s a common criticism of chatbots and other forms of personal financial management services that they don’t do much more than give customers different entry points to their transactional histories — does anyone really need five ways to check their checking account balances? The same can be said of fintech more broadly — that most new offerings are front-end solutions, but it’s the back-end of banking that needs an upgrade.

Banks surveyed by Celent, however, indicated they aren’t expecting more than 30 percent of their front or back office operations to be handled by AI.

“Erica won’t necessarily create a conversation from scratch. We have to help her understand: when customers say this, here’s what she needs to do. Theres a certain amount of back end work we can do to identify what conversation customers want to have with her and if she’s not prepared to deliver, we need to make sure the right conversation would be in those situations.”

How Kasisto avoids the financial chatbot fatigue

There’s an entire financial chatbot ecosystem emerging in the artificial intelligence space.

Chatbots, designed to simulate conversations with human users, have existed for a long time. Now, with so many millennial consumers who prefer digital interactions for accessing and managing their financial services, chatbot popularity has erupted.

Kasisto, which calls its MyKAI chatbot a “Siri for financial services,” is one of the most well-funded, having just closed a $9.2 million Series A round in January. It gives consumers a conversational platform over which they can ask about their bank account activity and allows them to link their Venmo or Facebook Messenger accounts to make payments initiated through KAI. It also works with banks to allow them to create their own conversational experiences with customers. The company was part of the inaugural class of Wells Fargo’s accelerator program and was originally a spin-off venture of SRI International. Incidentally, Siri was too before it was acquired by Apple in 2010.

Digiday spoke with Dror Oren, cofounder and VP of product at Kasisto, about how it separates itself from chatbot hype, how it’s adapted to technology changes in banking, the company’s long and short term goals.

How do you avoid bot fatigue?
The way we view ourselves isn’t as a bot company. We’re a conversational AI platform. We enable conversations on different channels, which can be chatbots on messaging but can also be conversations on mobile applications, on the web, through Alexa – we’re across channels.

The second difference is that we’re a platform. Yes, we have our own MyKAI bot but we really enable banks to build their own bots.

The third thing is, not all bots are created equal. There are dumb bots and smart bots. It’s easy to build a bot that demos well, it’s hard to build a bot that answers the questions you want and can deploy in an enterprise environment where you can train, retrain, scale, add more capabilities and have it run in a banking environment.

How has Kasisto’s vision of itself changed since 2013 given how much fintech and banking have changed in that time?
The value proposition around conversation for finance and redefining the way people interact with banks has been a consistent focus from the very beginning. When we were raising money, no one believed conversation would be the modality people interact with. Now we don’t need to convince people about conversation, but the conversation is a little different. Voice had higher appeal back in the day; we see it less today. Now the focus seems to be more on written conversation — messaging and chat interactions.

We’re also seeing an extension of use cases. It used to be about proving the ROI — how you justify the deployment of these systems. Now it’s about customer support messaging and banks are looking at more opportunities to extend their reach.

Won’t voice make a comeback?
We don’t see much demand coming from the market. One exception would be Alexa and the success of Echo. We’re piloting an integration with them so we know there are interesting use cases there. I don’t think standalone speech solutions add much value to your bank applications but adding a conversational element to devices that already do speech is probably something we’ll see develop.

It seems like every fintech company is touting an AI component now.
Yes, AI has become a buzzword. It’s everywhere but not clear where and in what contexts. We’re using AI in training our models, training a system that doesn’t already know the banking system. We’re also using it in run time; you ask a question to the bot and the system decides what the right answer is, what the “intent” is, what it is you’re really trying to do when you ask “how much have I spent on Uber between March and August?”

What is the future of MyKAI?
In the next couple years, we will see live deployments with banks solving simple but real problems around customer support, helping you understand your transactions, personal finance management; but also actual actions, like making payments, asking about reducing a fee, buying overdraft protection. Every bank and company will find their own valuable use cases and will double down on those. We’re lucky to be live early in the process because we already have a lot of data to help us look at what people are asking for and doubling down on those things.

How much does customer trust or distrust play into the evolution of the chatbot ecosystem?
There’s no real reason when you apply for loan or mortgage you’d wouldn’t want to with a bot. It’s only a matter of filling out a bunch of forms but right now people don’t do it. As time goes by we’ll see trust move to virtual conversations and away from live conversations.

And channel maturation?
We think a conversation can be cross-channel and blur the lines of how you think of the channel. If a user is going to a bank’s website on their mobile is that the mobile channel or the web? At the end of the day the consumers don’t care [about the channel]. They want a consistent experience across all the touchpoints.

How banks are using Watson

Despite banks’ simultaneous excitement and fear of artificial intelligence as perhaps one of the most transformative technologies for their business, they’ve been quieter about IBM’s Watson.

Watson is a cognitive technology and super-computer comprising AI that “learns” how to draw conclusions from data, natural language understanding – which allows it to read and understand unstructured data, like social media posts and digital photos – and a search engine that can comb through millions of data points in seconds.

It holds enormous promise in the long term for banks, who hold troves of customer data they’re constantly studying and using to create better customer experiences as well as improve operational efficiencies.

“What a Watson could deliver to banks would be tools to ensure sales people are selling the right things to the right people at the right time,” said Ryan Gilbert, a partner at Propel Ventures. “Wells Fargo probably wouldn’t have an eight financial product sales challenge if it had a Watson,” because Watson wouldn’t have allowed such rules to be set for employees to follow.

In the short term, these kinds of improvements will continue to manifest in digital banking chat bots, digital “personalities” programmed to be able to carry on a conversation with a customer. Banks are currently using other AI solutions for these experiments.

Here is how Watson has been used by banks so far.

Regulatory compliance
In November IBM bought Promontory Financial Group, an extremely influential strategy, risk management and regulatory compliance consulting firm in the financial services industry. In doing so, IBM hired its professionals — ex-regulators and former financial services executives — to teach Watson how to address banks’ compliance issues and ultimately create an AI capability that can sort through all the data banks collect to find problems and create solutions for critical needs around financial risk modeling, surveillance and insider threat, and anti-money laundering and Know Your Customer rules.

This marks the initiation of Watson’s move into banking, Gilbert said. It’s not clear how the Trump administration will move on the supposed unraveling of Dodd-Frank, the financial reform bill put in place after the financial crisis. However it moves, it won’t take away from the importance of regulatory compliance and the emerging so called regtech industry.

“If the administration does untangle a majority of the prior administration’s regulations there’s going to be a huge industry around compliance, a lawyer stream and it’ll be a compliance officer’s nightmare,” Gilbert said. “What better than an AI-powered system to gather data and get it all figured out?”

Empathetic bots
Royal Bank of Scotland is developing a chat bot called Luvo to answer customers’ questions in near-real time. Luvo uses IBM Watson Conversation, a cloud-based cognitive tool, which means computing systems learn as information changes or needs evolve. The service was made available in December to 10% of its banking customers through its web chat service and is still in the testing phase. By answering more basic customer questions, Luvo allows RBS advisers to devote more of their time to customers with more complex inqueries. It responds to customers to the extent that it can; if it’s too complicated then Luvo can pass them onto an adviser.

In the future, RBS plans to employ Watson Alchemy Language capability, which would help Luvo better understand customer sentiment – happiness, sadness, frustration – and change its tone and actions accordingly.

For example, it would be able to sense the difference between a customer needing to replace a lost card versus a stolen card. The latter can be a more emotional experience, and Luvo would probably pass the customer onto a human adviser. If someone simply can’t find a bank card, that’s something Luvo could provide some information about quickly.

 

Military separation advice
USAA customers, many of whom are current and former military members, can ask Watson questions and seek advice on transitioning back to civilian life on the bank’s website.

The Watson Engagement Advisor answers questions related to military separation on topics like job searching, home purchasing, military benefits and more. For example: “Can I be in the reserve and collect veterans compensation benefits?” or “How do I make the most of the Post-9/11 GI Bill?” This requires that Watson comb through volumes of USAA’s business data to feedback answers to member’s inquiries.

Wealth management advice
Australia’s ANZ Group has perhaps been Watson’s highest profile banking user. The bank employed Watson Engagement Advisor for its wealth management offerings. ANZ staff — advisors, product experts, legal and compliance staff and customer service people — feed documents and data to the supercomputer about the bank’s products, including their latest terms and conditions.

The technology is meant to help personnel assist customers with deeper insights and at a faster pace, but also employs the Ask Watson feature — the same used by USAA — to give customers feedback to guide their purchase decisions and troubleshoot their problems.

Personalized banking
Citigroup has a long working history with IBM to bring information technology into financial services but it was just a few years go that it brought Watson into its business to explore ways to advance analyze customer needs, improve customer interactions and process vast amount of financial, economic and client data.

At the time, Citi said using Watson’s content analyzing and learning capabilities would help it deliver more simplified banking services, intuitive branch experiences and personalized banking.

Why banking’s ‘omnichannel’ dreams haven’t become reality

For years banks have been talking a lot about executing an “omnichannel strategy,” which is supposed to help them learn more about their customers by giving them a greater view into their needs and behavior through more channels – the mobile device, the tablet, the computer. But that dream hasn’t yet become a reality.

The reasons, as usual, are in the mundane details — and in the difficulties in executing sensible strategies within large, hidebound organizations while keeping up with new complexities that inevitably arise from tech advances.

For instance, an omnichannel strategy requires banks to be able to integrate each of the channels into a single, quality experience. But most haven’t gotten there. According to a report this week by researcher and consultancy Celent, half of the 112 institutions it surveyed haven’t even begun “substantive” efforts on their omnichannel delivery and just one in 10 institutions is actually executing a strategy.

“There’s a huge disconnect,” said Bob Meara, senior analyst at Celent. “Everyone agrees omnichannel is important but they haven’t actively executed.”

One reason is that artificial intelligence is driving a proliferation of new channels – like Alexa or connected cars – that make it impossible to build experiences for individual channels well and in a scalable way. So while many banks still struggle to perfect their mobile strategy, the ones that are nailing the “omnichannel” idea are now having to move along pretty quickly anyway as new technologies and therefore, new experiences, emerge, said Meriah Garrett, chief design officer at USAA.

“Our members just expect us to be there wherever they are,” she said. “That’s not always in this pure mechanism of traditional channels as we once thought of them – mobile, web, voice, physical. Those things are blurring together at such a fast rate.”

To keep up with the constant change, banks need to implement AI into their interactions and services, and that’s how the “channels” expand beyond what people traditionally consider a channel to experiences like a Facebook Messenger conversation or a mortgage profile in Zillow.

“It becomes less and less about any individual’s channel and more about different distributions of experiences that aren’t necessarily owned properties anymore — that’s the part we as an industry have not even reached yet.”

Most financial institutions have invested a significant amount on the front end of their banking portals – the parts that interact with customers. Some might say they’ve over-invested in that experiences when they should be pouring more into the middle- and back-end – the areas that actually connect with other experiences, other parts of the business and improve seemingly boring efficiencies that actually make a world of difference to the customer.

For example, getting approval on a personal loan has traditionally been about a 72-hour process – unheard of for customers living in an on-demand world where you can get a car at the tap of a button. That kind of thinking is what makes digital lending startups like Prosper, Avant or Kabbage so attractive when they advertise decisions in minutes. If banks invested more in this stage of the experience, customers wouldn’t just be happier, they would probably engage more consistently.

It’s how Amazon rose to dominate retail. Jeffrey Brown, global banking and financial services leader at consulting firm Genpact, uses Amazon as an example when advising his bank clients, he said.

“You want to create Amazon Prime in your banking experience for your customers,” he said. “You wouldn’t use Amazon if it took 10 days to turn around a delivery. You get the Best Buy experience when you do your banking.”

Using mobile or online banking as a reference for account balances and activity is more common than ever. But actually executing on more complex things like credit applications is still a sticky spot for banks and their customers. The user interface of complex activity may be enjoyable, but the parts of the experience that follow need to meet the same standard to keep customer satisfaction levels high.

“Just like they shifted from retailers and other people who couldn’t get them the goods they wanted quick enough and thats why amazon took share.

“Being able to actually deliver, execute or fund is going to move market share over the next 12-24 months, Brown said. “People want speed. They shifted from retailers and other people who couldn’t get them the goods they wanted quickly enough and thats why Amazon took share.”

By investing in biometrics and AI, Wells Fargo is eyeing a move into voice payments

Wells Fargo is working on a voice-first payment capability it could soon make available for consumers.

That would go beyond where the rest of industry is at with voice interface, which, for the most part, has not yet advanced beyond basic requests like, “How much money do I have in my checking account?”

In 2015 for example USAA launched Nuance’s virtual assistant, Nina, on its website. Bank of America expects to launch Erica later this year as a virtual assistant integrated into the mobile banking app. Capital One has an Alexa skill. Customers can turn to their virtual financial assistant for basic day-to-day functions, like checking account balances and credit scores of scheduling bill payments.

But all of these things just scratch the surface, said Steve Ellis, head of Wells Fargo’s innovation group, who teased the bank’s own forthcoming voice-controlled payments offering.

“These digital assistants like Siri or Alexa — these things are just starting,” Ellis said. “There’s a really big future here for how our customers interact with us. We are starting to do proofs of concept with information exchange… but the idea of moving money from a fund to someone through a peer-to-peer payments system — that’s coming.”

He didn’t specify when it would introduce that functionality but hinted it would be “a shorter time frame than three to five years.” In the meantime, the bank is exploring how it uses the artificial intelligence that provides conversational banking abilities, and is bulking its biometric authentication practices necessary to nail mobile voice banking.

People are embracing the idea of conversing with digital assistants to exchange information and even make Amazon shopping purchases. Apps like Uber have raised customers’ standards for experiences that are fast, seamless and secure. Owners of smartphone devices are getting used to authenticating using their fingerprint, even if it’s just to unlock their phones; Apple claims a user will do this 80 times a day.

Banks, however, have yet to make the customer experience completely password- and PIN-less.

“We have about 5 billion-plus interactions with our customers every year and every one of them starts with authentication,” Ellis said. “If you can’t get that right, you can’t do anything else, so that’s always our first focus.”

Wells Fargo is no stranger to biometrics. Ellis himself co-launched its startup accelerator in 2014 when EyeVerify approached it with an idea to develop eye-recognition technology for security and identification purposes. Today, Wells uses EyeVerify’s eyeprint verification for its commercial banking customers.

Chatbot provider Kasisto, which Ellis described as a Siri for financial services, was selected for the same inaugural accelerator class. Whether Kasisto has the success of EyeVerify remains to be proven but last month the chatbot startup raised $9.2 million to expand its virtual assistant offerings. Ellis said Wells is “very close” to rolling out some of those virtual assistant capabilities to select employees and customers on a test-and-learn basis.

Besides the biometric factor, Wells and its peers need to work on the quality of the AI and its speech recognition as well as its ability to truly understand and process what a customer is saying on an emotional level, said Ron Shevlin, director of research at Cornerstone Advisors and author of Smarter Bank.

“There’s a lot of potential for [voice banking] but I’m afraid we’re conflating the voice interface with the AI capabilities needed to interact in a high quality, whether it’s providing service or advice, he said. “If it’s simple types of interaction then the bloom is off the rose… There’s no economic impact, no greater levels of customer satisfaction you’ve just created one more way for someone to check their account balance or account fraud or maybe pay a bill.”

Ellis identified three markers on the road to a voice-first future: information exchange, funds transfer and personalized advice. Most banks are at the beginning. Wells Fargo hopes to be an early mover into the second stage.

That was part of the reason for refocusing its efforts inside the Innovation Group. This week it announced it would be dedicated to AI as well as payments and application platform interfaces.

“AI is a baby step right now,” Ellis said. “It’s going to really adjust the way people think of how they use their phone get information and actually do things.”

Why robo-advisers are looking to former magazine editors for the human touch

Robo-advisers Wealthsimple and Ellevest believe in the human touch after all.

Both have plucked editors from top publications in order to personalize the often-dry world of investment advice by focusing on lifestyle matters. Ellevest hired chief design officer Melissa Cullens, who was an independent design strategist that Vogue.com recruited to lead the re-design of its website. Canada-based Wealthsimple, which expanded to the U.S. last week, hired Devin Friedman, a former GQ editorial director, as brand editor to lead its section that features interviews with people about the role money has played in their lives.

“The human side is something I don’t think a lot of people in our space have done a great job of, and it’s one where we excel,” said Wealthsimple chief product officer Rudy Adler. “It’s easier to do when you have a great voice; you’re coming at them with humor and not boring them with dull finance articles. We’re trying to find an emotional way in.”

At Ellevest, the investment platform for women launched by Wall Street vet Sallie Krawcheck, Cullens focused on the role money plays in helping people feel safe and express their values, like having control over their lives. That thinking was expressed in the design of the platform but also the user experience, including how many steps users have to take in the on-boarding process, what information they hand over and what the form field experience is like.

For example, similar services often ask questions about a user’s investment experience and risk preferences, and they’re usually laden with jargon and can take at least 10 minutes. Cullens designed the Ellevest on-boarding experience to include the client’s life goals, not just financial goals, and streamlined the process.

“We wanted to try to create an interface and experience that gave her the reins she was already taking and make investing work for her instead of making her learn more, work harder or be better to fit into the mold of a system that ultimately has excluded women,” Cullens said.

Robo-advisers, which dole out artificial intelligence-driven investment advice through a website or app, have been one of the fastest-growing parts of the U.S. fintech market. But that growth has leveled off as banks and other traditional financial institutions have piled into the space, creating new competition for customers.

Traditional financial services have mostly targeted high-net-worth individuals. But fintech depends on being able to relate to young people who are new to investing and may have lived through the last recession and distrust traditional finance. That’s where people with luxury brand experience can help robo-advisers differentiate.

“In the early days, no one thought about client experience; they just thought that if you log in, then you can make a transfer,” said April Rudin, chief executive of wealth management marketing firm The Rudin Group. “By bringing talent that has experience in creating a luxury experience, the idea is [fintechs] will take their functionality and give it client experience.”

AI becoming just another tool in the trader toolbox

In the collective mind, the trading floor is often depicted as a chaotic and noisy place where traders shout out orders at the top of their lungs. The reality is much much quieter. Where humans used to shout, bits and bytes now move silently.

The process started when trading shifted from manual, voice-based actions to computers. Then it developed further when automated order routing was added, saving time and resources. Then, automated trading systems rose in popularity, responsible for about 75 percent of today’s market volume.

In recent years, with the advent of cheap computational power, the next step of automation is trading automation driven by AI or machine learning. It is here that popular media propagates fears about the ultimate takeover of machines.

“It feels like some of the buzz and the hype has died down, which is an interesting stage in the lifecycle,” said Tom Doris, CEO of OTAS Technologies, a market analytics and trader intelligence company. Now, he adds, we can get back to looking at the problems that technologies like these are supposed to solve.

A typical trader, explains Doris, has too many orders in his queue. Faced with this, he will start processing them from the top and work his way down to the bottom. He might linger a little bit more on those that are more volatile or require more attention, but it might be hard to spot those among the noise.

“The market is generally pretty boring,” Doris said. “If you have 100 orders, 95 of them are perfectly ordinary. It is all very predictable. Your task as a trader is to find those 5 percent where something unusual is happening.” OTAS’ technology is able to create a predictive model of how a stock should behave, and alerts the trader when unexpected information changes a stock’s behavior. The technology can be plugged directly into an Execution Management System, so that the trader can act on those alerts instantly.

The trend over the last couple of years was towards increased automation, attempting to take humans out of the loop. Now, it’s understood that there is a limit to what a trader can do systematically.

“You want humans looking at situations where there is a human story going on,” Doris explained.  For example, when a stock starts to rally because of aggressive buying, only a human with a good understanding of the risk landscape and the company’s story can discern if this is informed flow or a trend that may revert.

There have been several hedge funds priding themselves in the use of AI software to guide their decision making, including Bridgewater Associates, Renaissance Technologies, D.E. Shaw, and Two Sigma. Many more firms describe themselves as “systematic”, meaning they base their decision making on computer models, which might not be driven by AI.

Perhaps the most common approach to AI in investing is the use of natural language processing to be able to make sense of unstructured market data and the use of neural networks to identify patterns, relationships and hidden trends.

AI is seen now more as another tool in the toolbox of traders, rather than a magic bullet, Doris concluded.

WTF is cognitive banking?

Your next bank might be Skynet, if IBM has any say in it.

Though many companies offer AI solutions for financial institutions, IBM is championing the use of cognitive computing in banking, publicizing the term as a new paradigm.

What is cognitive computing and how does it apply to banking?

Based on machine learning, natural language processing, and human interface technologies, cognitive computing systems can learn as information changes and requirements evolve, and easily interact with users, other devices and other data sources. In contrast to traditional computing models which tabulate and calculate based on preconfigured rules and programs, cognitive systems can handle situations that are dynamic and information rich.

When applied to banking, cognitive computing can offer a wide set of benefits to both banks and customers.

What does a fully cognitive bank look like?

Imagine banking was as simple as a Google search. Instead of surfing multiple pages on your bank’s app, you type or talk to a  single input box: “I lost my card.”  A quick chat with a rep you didn’t even realize was not human and the new card is on its way. Compare this to the current frustration of clicking multiple links to find the right phone number, and one can easily see the benefits.

Banks can also leverage machine learning to predict financial needs and proactively suggest to a customer a personalized loan or to transfer money to a saving accounts ahead of a coming purchase. Leveraging a lifetime of customer data, banks will be able to truly personalize the services they offers each client, at scale. A seamless omni-channel experience, often conversational, can turn banks into a trusted advisor who is available when we need it.

Since a cognitive system learns and improves with every iteration, a virtuous cycle of customer satisfaction is formed.

Are banks adopting the new paradigm?

If you banked in the last week, you probably know the answer.

Generally speaking, banks are still testing the waters when it comes to cognitive banking. In a 2016 survey conducted by IBM, just 11 percent of bank executives reported they have adopted cognitive technology. 58 percent named improving operational efficiency as their most important strategic priority right now, which might explain the low adoption rate. Banks are generally focused on cost reducing activities and do not make needed IT investments.

In the same survey, 49 percent cited [the rather superficial outcome of] operational efficiency as the main benefit of cognitive computing, indicating bankers are a bit aloof to the transformative potential of it.