On Thursday 14th September I attended the AIR Events conference in London. You can download the programme PDF here.
It was fascinating: good mix of speakers, well curated, well chaired, and with decent food. I took loads of notes and am aiming to reproduce the more legible ones for you. Here are the first two, Chris Corrado Group COO and CIO at the London Stock Exchange, and Pete Trainor from BIMA (British Interactive Media Association):
Session 1: Helping money make the world go round: how AI underpins safety, security, data quality and profitable operations for London Stock Exchange
Chris Corrado, Group Chief Operating Officer and Chief Information Officer, London Stock Exchange Group (‘LSEG’)
Summary: The LSEG is heavily investing in AI for, amongst other things, market surveillance and cyber security. They’re using IBM Watson and have purchased Mergent a data intelligence company.
In Chris’ view AI value creation means:
- projecting the future
- producing a product or service – one that the customer wants and is higher quality & lower cost
- promoting an idea
- providing a better experience
The LSEG has a list of things they can do with AI in a business (will they share the slides?), here are a couple:
- Market Surveillance
The LSEG runs 3 exchanges, they want to run them fairly. Conventional fraud detection is static – LSEG wanted to be smarter and for their people to be less stressed and more productive. Much of this was about setting rules and boundaries but a lot of guesswork! So they are using technology to:
- reduce false positives
- pick up false negatives
- detect anomalies
- ingest more data
- market analysis / news flow analysis
They put data from the stock exchange into IBM Watson. They detected new anomalies, partnered with a software company that the LSEG already used and put it into production – they now have a better system for detecting anomalies, in other words, fraud.
- Cyber security
- LSEG has significantly stepped up investment in cyber security: they’re living in ‘a bad neighbourhood’ and it’s getting worse
- Already running anti-malware and are investing to make it better – AI based
Before going to LSEG Chris spent more than £1Billion in 3 years at Morgan Stanley. And he told them they were a data company – got thrown out the room the first time he said this. The LSEG used to depend on other business to get their data. They now provide it and ‘manufacture it’ themselves – Mergent (400 people and a recent LSEG acquisition) – have developed an AI platform to ingest data from public and purchased reports: they clean and validate it in a much more productive way than before.
Success factors to date
- Business teams and tech teams working closely together
- Healthy partnering environment: “we want to solve a problem not sell you a product”
- Mature testing processes: an acceptance that not all projects will succeed, proof of concept phase is normally 8 weeks and then continue or retire it
- Healthy experimental, entrepreneurial working environment
- The future is augmented intelligence – not ‘artificial’ any more, e.g. Amazon Alexa
- Build a cognitive business: derive meaning from all forms of multi-structured data
Session 2: What is AI and how is it going to break my business?
Pete Trainor, BIMA, the BIMA AI think tank, co-founder US AI Ltd
Summary: Pete rattled through a critique of technology and AI, and how people have become utterly dependent on their smartphones. He then introduced us to ‘Su’, a ‘question and answer’ app that addresses the issue of male suicide.
I’m not sure how many delegates spotted the irony of an app to addresse the problems of smartphone dependency…
Here’s what Pete told us:
Technology gets in the way, eroding the best part of humanity… more people are looking at their feet not the stars… stop staring at screens and start having conversations again.
Before we were users, we were people; we need to empower AI by approaching it in a human-focused way. You can plot a user journey but actually, you know nothing about them.
Smart use of data could give us help before we need it. Behaviour is a function of your personality and your environment. There are 4 parts to your digital soul: the app, the notification, the Operating System, the data.
Ee can now track how people feel not just way they do: cognitive biometrics. AI needs a sense of purpose – EQ not IQ
Case study: meet Su
- 5 guys spend 80% of their time in just 5 apps
- They all suffer from a virus
- In 2015 they all took their own lives
- In response to this, Pete’s team used data analytics, natural language processing, emotional analytics and telematics to build an app called Su www.hisu.chat
Pete told us “Su is not a piece of software but a member of our team; we’re teaching her to ask lots & lots of questions”.
One of the users told them: “this is better than talking to a person, I don’t feel comfortable talking to a person about these things”.
However, this is surveillance technology – we need to balance how we use it.