The Fourth Industrial Revolution, in which augmented human intelligence is creating a confluence of the physical, digital, and biological realms (as defined by the World Economic Forum’s Klaus Schwab), is well under way. Computing systems powered by high-performance cloud and quantum computing, blockchain (distributed ledger running on peer-to-peer networks), big data, Internet of Things, bio-hacking, robotics, and advanced analytics enable the financial industry to move from a “coding” to a “learning” world.

In the coding world, the aim was to understand what businesses wanted to do and produce programs that executed those requests. The learning world is characterized by systems with three essential elements: knowledge, skills, and experience. Computing systems are now being produced that learn from and with people—that is, they augment people rather than replace them. This is called augmented intelligence (AI), and nowhere is AI clearer than in the financial services industry.

In a recent Baker McKenzie survey, the financial sector identified three main areas where AI will be applied in the next three years: risk assessment, financial analyses, and investments or portfolio management. Global asset managers and hedge funds like BlackRock, AHL, Euclidean, RBC, Renaissance Technologies, Two Sigma, and Bridgewater Associates are pouring money into developing next-generation investment systems that can outperform any human. Man Group, a $103.5-billion firm in London, England, already devotes about US$13 billion to several hedge funds using machine learning. “In 10 years, AI will play a role in everything Man does, from executing trades to helping pick securities at the firm’s discretionary unit,” said the firm’s CEO, Luke Ellis, in an interview with Bloomberg. Orbital Insight, a start-up backed by Sequoia Capital, is using big-data analysis of satellite imagery with AI to give investors a more accurate and timely view of the global economy. For example, photos showing the depth of shadows cast by floating lids of giant oil tanks in China that the world’s largest energy consumer may have stored more crude than official government estimates.

IBM’s AI platform—Watson & associated developers + data ecosystem—is already at work in the world, serving more than a billion people. Watson is being used to expand expertise and improve decision-making in a variety of areas: health care, financial services, natural resources, law, retail, and education. Today, hundreds of clients and partners across 36 countries, and more than 17 industries, have active projects under way with Watson.

In financial services, IBM recently publicly announced new client engagements with Thomson Reuters, Swiss Re, and KPMG to adopt Watson to help transform their business solutions. IBM has usually been engaged in three primary domains: client experience and engagement, investment management and advisory services, and risk and compliance. The following are some examples of these engagements.

Client experience and engagement

AI allows for a completely new and personalized experience in which the more the system experiences, the more it learns and the better it performs. One large bank is developing cognitive insight and marketing capabilities for the small business advisor to provide value-added advice to small business clients, which will, in turn, translate into an increase in revenue and customer retention. The primary focus is to build deep learning models using interaction data that will look for time-sensitive cross-sell and up-sell, predicting churn and generating leads from retail banking clients. Another expected outcome is the improvement of the consistency of services delivered, no matter the channel, the geography, or the person providing the experience.

Investment management and advisory services

San Francisco-based EquBot has launched the first-ever ETF to use IBM’s Watson. The AI Powered Equity ETF (NASDAQ: AIEQ) attempts to mimic an army of equity research analysts working around the clock. The fund uses AI to scan more than 6,000 U.S. publicly traded companies each day to create a diversified fund. Watson parses regulatory filings, more than a million news stories, company management profiles, sentiment gauges, and financial models to create a portfolio of 30 to 70 stocks. AI systems can help analysts, advisors, and clients understand and make faster, more accurate investment decisions.

Risk and compliance

Since 2008, global banks have paid more than US$320 billion in regulatory fines and penalties. Some AI systems can ingest new regulations as they’re created and review 800 million pages of text per second, providing decision support on issues as diverse as regulatory reporting requirements for derivatives and consumer protection for mortgages. Watson streamlines the identification of potential regulatory obligations, reduces the time and costs of compliance, and enables sustainable management of controls through an easy-to-use dashboard.

Lessons learned

According to a recent Expert Insight study by IBM’s Institute for Business Value, much can be learned from the experiences of first movers who have used AI to become more customer-centric.

1. Start with a holistic AI strategy and business case with an enterprise-wide governance framework. Identify opportunities that will impact customers, and get commitment from senior management to follow through.
2. Rethink and redesign the entire user experience. Use design thinking, a framework for AI learning systems, and actions that promote a focus on user outcomes. The benefits of design thinking include helping employees to develop new ideas and prototypes, along with the powerful words and imagery to describe those new concepts to peers and customers.
   
3. While siloed data sources must be prioritized, aggregating as many of these sources as possible is foundational. More data can lead to greater personalized solutions such as obtaining a mortgage, reallocating a portfolio, or evaluating economic factors to make investment decisions.
4. Use a highly integrated, well-orchestrated AI system, because there’s no assurance that disparate components will scale and work in a 24/7 production environment.
5. Give the AI solution a name and personality to allow users to form an emotional bond and empathize with it. Consider the gender of your AI, and how formal it is when communicating with users.
6. Remember that AI systems are meant to augment human performance—not replace it. Keep the use, and potential, of AI in perspective at all levels throughout your organization.

Looking ahead

On Dec. 14, 2017, the world took a leap forward toward quantum computing readiness. A dozen international organizations representing Fortune 500 companies, academia, and government joined the newly minted IBM Q Network. While it’s true that the promise of quantum computers is likely still a decade or more away, IBM Q systems are the most advanced superconducting universal quantum computers available at the moment. With these systems, scientists will continue to be able to push the quantum computing field toward the initial demonstrations of “quantum advantage,” with applications in chemistry, optimization, and machine learning, and which will eventually give rise to the field of quantum AI. When quantum AI becomes available, only change and relationship management will continue to matter for wealth and pension managers, making everything else obsolete.

Hear Pavel Abdur-Rahman on the applications of AI and its implications for pension and investments at the 10TH ANNUAL SPRING PENSION CONFERENCE on April 5, 2018, in Toronto.