Artificial intelligence (AI) will one day become so commonplace that it will fade into the background, according to attendees at a roundtable held by Realization Group and Software AG on 7 June 2017.
AI is already embedded in tools, including anti-fraud systems, chatbots and security software but, with regard to financial services, many impediments to operationalizing AI must be addressed before ubiquity can be achieved.
The key takeaways were these:
- Firms must remain realistic. Identify small but appropriate use cases and build out from them. The democratizing of the tools and technology that underpins AI has allowed a critical mass of people to develop these tools and made them useful for financial services firms. Nevertheless a cultural change is needed so avoid a revolution. Ensure youhave sufficient quality data in order to train your models.
- Centralized management of data is crucial to enabling trust in AI systems, through quality assurance and maintenance.
- A standardized platform and methodology for deploying AI logic across the enterprise with appropriate monitoring and governance to avoid developing more complexity in the IT infrastructure.
- Cloud can provide elasticity to handle large data sets but it has challenges. Some enterprises resort to building private cloud instances within public cloud offerings in order to overcome concerns about data security.
- Regulation, specifically transparency, can also be a barrier to adoption of AI. If a system were to suffer from reduced accuracy, identifying that change without full transparency over the processing could be difficult. Black box decision-making is out of favour.
- Attract and retain the best people through engagement. This will drive the cultural change. Attracting these people is challenging, however, for example where there are barriers to engaging in open source projects. Working with third parties who can offer the right rewards can create this access to talent.
- The potential upsides of AI are enormous. There is an opportunity to pick up whole new ranges of activity that were never conceived before.
The room agreed that, when setting up an AI project, keeping one’s feet on the ground has to be balanced with the excitement needed for getting funding and the reality of delivery. Unfortunately, projects which promise the moon are often given the green light, while more mundane yet realistic gains, such as improving administrative efficiency, are not.
Firms need to plan a route to get to the AI-enabled future, with awareness of the current challenges and strategies for addressing them. The underlying architecture needs to be put in place at this stage, so that analytics can be run across data sets and data can be made accessible. Simple example AI applications, such as password resets, are good early stage developments that develop expertise.
The conclusion? There are challenges to making AI work alongside financial markets, but the opportunities far outweigh any of the hurdles. Watch the webinar, Naturally Intelligent - Deploying AI Systems into a Live Environment, here.