Once, we were very confident that we’d have true artificial intelligence (AI) by the end of the 1980s (remember the hype around expert systems?). Also, the first AI conference was held in 1956 - and the participants thought they’d see it in their working lifetimes.
The overblown promises of the AI crowd have been quietly forgotten. Or, have they? An insightful definition of AI is, ‘everything a computer can't do yet’.
What that remark crystallises is the remarkable quiet advance computing intelligence has been making over the years. Many complex or abstract capabilities that have long been thought to be exclusive to the human mind are no longer so. After all, there are computers that can beat world chess champions, play TV games shows to a winning level and accurately recognise faces or drive cars.
Machines have massively expanded the range of activities in which they can match - possibly on their best day, even - outperform homosapiens.
All that’s been made possible via quantum leaps in available computing power, combined with advances in learning and pattern recognition algorithms. As a result, computers can now ‘see,’ ‘listen,’ ‘read,’ and ‘write.’ Often in a very constrained fashion, but there is undeniable understanding on their part, in the sense that they can interpret natural language and react accordingly.
AI has slipped out of the lab, in fact, and sits on all sorts of applications and devices that you wouldn’t really have considered - from AI in computer gaming to Skype, which can now translate human conversations in real time.
Speech recognition and computer dialogue is now a commonplace, via applications like Siri, Google Now or Cortana. So-say sentiment analysis allows programs called agents that are springing up everywhere to determine the emotional state of their interlocutors - and react accordingly.
We’re starting to get used to interactions with smart agents, which are getting slicker, more natural-seeming and more pervasive. That would have been seen as a proof of concept back in 1956! But computers can now do more than interact with humans. Smart (sometimes labelled ‘cognitive’ technologies), in their various forms, are now shaping many businesses and industries.
Virtual assistants provide technical assistance to experts in aeronautics or oil-platforms to perform complex constructions and repairs, while embedded smart systems allow cars to self-drive many miles without accidents.
Robots can be taught by, literally, showing them the actions to perform (no programming required). And IBM’s Watson - the successor to the system that beat Garry Kasparov and which did so well in the game show Jeopardy - has contributed to cancer research by improving diagnosis accuracy, as well as revealing new mechanisms of the disease.
Watson helps chefs explore new flavours and create new dishes, while a system from a firm called IPsoft, Amelia, a smart agent with a convincing avatar, can read and understand text, follow processes, solve problems and learn from experience. Notably, she understands implied, not just stated, meanings, and improves her performance by hearing humans deal with questions she can't yet answer.
Amelia can digest an oil-well centrifugal-pump manual in 30 seconds - and give instructions for repairs, do the job of a call-centre operator, a mortgage or insurance agent, even a medical assistant, all with virtually no human help.
To achieve these feats, the software is never set up to cover every possible situation, as not only is this an impossible task, but is also pointless in many situations (how can you write a procedure if you don’t know the solution to the problem - e.g. curing cancer?). Instead of brute-force calculations, pre-programmed responses and keyword look-up, most of these systems are deigned to be highly flexible, able to learn to recognise patterns and infer rules from examples.
This learning component is key to making cognitive technologies work in real environments. Learning agents don’t need to be reprogrammed each time a new situation arises. They naturally improve and expand their capabilities with time. Which brings us to a key question: could this also start happening in industries such as financial services?
Levelling up financial services
We think yes - as the sector naturally lends itself to the use of cognitive technologies. The complexity of the financial markets, the vast amount of data, and the need for automation and better customer experience make cognitive technologies a convincing solution in a wide variety of situations.
In risk management and compliance, for example, smart agents could easily evaluate all cases against approved policies and guidelines and understand the complexities of risk exposure.
Financial and market analysis could be made more insightful through the analysis of vast amounts of information. And, in sharp contrast to traditional analytics, smart agents are more than happy, thanks to that design philosophy mentioned, with open-ended questions, detecting key trends and variables.
Today, in wealth management, relationship managers advise their clients by analysing large volumes of complex data such as research reports, product information and customer profiles. How big a step is it before smart advisors could also provide cost-effective, personalised investment advice based on the ever-growing corpus of investment knowledge?
In fact, systems like Watson and Amelia are already used by top financial institutions. DBS Bank uses Watson to identify the needs of wealth management customers and offer better advice and determine optimal financial options; one of the biggest US banks uses Amelia to manage trading platforms and call centres.
There is also Kensho’s Warren which can assess how different securities are likely to react after the release of a market-moving piece of information and provides advice to investors. The developers of the system expect it will be able to answer more than 100 million distinct types of complex financial questions - sooner rather than later.
The conclusion seems inescapable. We have been too dismissive about the limitations of AI, and we are now at a point where cognitive technology capabilities are powerful and reliable enough to be deployed in complex business environments - including our own, the global financial services industry.
Financial services, like other industries, will benefit from these advances; with cognitive agents, banks and other financial institutions can readily improve their operations and services.
As with all moves to automate, cognitive systems will help reduce costs, improve processes and productivity and help banking operatives focus on their core business. And, as their capabilities expand, cognitive technologies will bring more and more value to the banks’ customers, for example, in terms of investment advisory.