There’s lots and lots of chat about AI, Artificial Intelligence, in banking using deep data analytics to augment our financial lifestyles. This is nothing new. After all, Spielberg made a film about it way back in 2001. What is new is the developments in AI that we’re seeing from firms like Google and IBM.
IBM’s Watson became famous for winning on Jeopardy! four years ago …
Watson (named after IBM’s founder, Thomas J. Watson) is a cognitive technology that processes information more like a human than a computer. The system is based upon an IBM supercomputer that combines AI with sophisticated analytical software for to provide a “question answering” machine.
The Watson supercomputer processes at a rate of 80 teraflops (trillion floating-point operations per second). To replicate (or surpass) a high-functioning human’s ability to answer questions, Watson accesses 90 servers with a combined data store of over 200 million pages of information, which it processes against six million logic rules. The device and its data are self-contained in a space that could accommodate 10 refrigerators.
Watson's key components include:
- Apache UIMA (Unstructured Information Management Architecture) frameworks, infrastructure and other elements required for the analysis of unstructured data.
- Apache's Hadoop, a free, Java-based programming framework that supports the processing of large data sets in a distributed computing environment.
- SUSE Enterprise Linux Server 11, the fastest available Power7 processor operating system.
- 2,880 processor cores.
- 15 terabytes of RAM.
- 500 gigabytes of preprocessed information.
- IBM'sDeepQA software, which is designed for information retrieval that incorporates natural language processing and machine learning.
Here’s how it works:
And, if you’re really interested, here’s a one-hour documentary on Watson:
Google meanwhile are using DeepMind, a London based company the search giant acquired in early 2014, to create their own AI program. DeepMind is now doing some amazing things, including creating a program that can beat humans at video games …
… and another that’s worked out what a cat is just by analysing YouTube videos.
The incredibly brainy Demis Hassabis, CEO of DeepMind Technologies, explains more about the implications of AI in this 16-minute lecture …
… and, if you’re really interested, here’s a 90 minute lecture from Stanford that tells you all you need to know about deep learning from Big Data analytics.
There are several other AI product developments out there too, including Microsoft’s Project ADAM (Active Directory Application Mode); Facebook’s Open Sourced Deep Learning tools; Apple’s evolution of Siri and the iOS recognition systems; and Amazon’s Machine Learning Service.
These are just the big guys as there are also hundreds of small guys doing neat stuff out there too.
So what’s the point of all this AI development? Well, in banking, it’s quite important as the technology giants are basically training enormous networks of machines to identify faces in photos, recognise the spoken word, and instantly translate conversations from one language to another. This means that not only can people talk with their bank when their bank is a machine, but the bank that is a machine can also immediately recognise the correct versus the fraudulent transaction.
I’ve already talked about this a while ago, when PayPal used deep learning to track fraudulent transactions, but there are many other applications of AI for banking. For example, several of the new payday lenders and credit firms are using real-time credit scoring analytics to calculate the credit worthiness of applicants. Equally, deep data analytics for marketing (effectiveness of campaigns), trading (to build predictive models of prices, volatilities etc), portfolio management (as a source of alpha) and risk management (to try and obtain better risk estimates) are all areas developing fast.
I was impressed with UBS earlier this year, who proudly told me that they run deep data analytics combined with machine learning non-stop on their clients’ portfolios of investment, to better advise each customer with specific and personalised services every day.
Equally, I was intrigued to hear DBS talking about using IBM’s Watson earlier this year. Similar to UBS, DBS are using deep data analytics to improve customer service and advice. Instead of spending more than two hours every day poring through market reports, DBS’ relationship managers use the time to meet with clients instead, armed with information from those reports distilled by supercomputer IBM Watson.
Certainly we are going to see more and more use of AI for everything from simpler user interfaces, improved customer experiences, automated fraud detection and massively personalised, proactive and predictive services. But where does all of this take us long-term?
Maybe ANZ’s partnership with IBM’s Watson can tell us …