Artificial Intelligence Set To Wow The Finance Sector's Traditional Intelligence

You may have heard of IBM Watson already, but in case you haven't: it's a cognitive computing capability that is commercially available. Named after IBM's first CEO, Watson beat two champions of the game-show Jeopardy back in 2011, and more recently has spent some time watching popular films, even making the astute observation that Han Solo is among the most neurotic of characters in the Star Wars universe, suggesting that underneath the cool exterior is perhaps a more complex constellation of emotions—I don't think he's wrong. (He also had opinions on the Harry Potter series).

But of course, plenty of people have plenty of plans for the intelligent robot outside of TV and film. Blockchain enthusiasts will be happy with the news that Watson could potentially enable regulators to easily find anomalies by having Watson scan huge volumes of past transactions or even help devices that use distributed ledgers self-diagnose errors. These applications are perhaps just the beginning, with Tim Hahn, IBM's chief architect in charge of Internet of Things security, suggesting that we're starting to approach "the kind of things we see in movies."

In effect, what Watson does is analyse unstructured data, i.e. data that wasn't produced specifically to fill in one field of one form: articles, books, tweets, quotations, laws, speeches, commentary, and every word I'm writing right now. IBM's website says that 80% of all data out there is unstructured, which Watson combs through using natural language processing (NLP). NLP means that he deconstructs every word, phrase, sentence, and paragraph, runs tests to statistically determine what the intended meaning is, and then can use that meaning for whatever analysis you fancy using it for. It was, therefore, only a matter of time before somebody plopped him down in front of the original Star Wars trilogy.

But naturally we're also interested in what Watson's intelligence could do closer to home, specifically in the audit business. Since Watson can read vast amounts of a company's unstructured data, he could help auditors identify anomalies, expand their ability to test those outliers, and generate evidence that will support or contradict an audit opinion. In short, patterns will be findable that were never findable before, and quickly too. I'm picturing Neo decoding the avalanche of green symbols in The Matrix, and I think that 1999 us would be proud (uh...