I’ve just been in a meeting with a project team developing an AI chat tool designed to help users search content and services more easily. They’re moving out of private Beta, with a cohort of users actively engaging and sending feedback.
I found myself reflecting on the team members driving this development and who else might be affected by it.
Many see AI as a way to expand capabilities and boost productivity for users and the organisations deploying these tools. But does the concept of “release” hide known risks that teams, eager to build, may be overlooking?
Bottlenecks downstream
It’s great to develop a powerful search tool that empowers users to explore an organisation’s services and products. But be aware of how this affects the teams responsible for delivering these products and services. They may not have adapted to the new demands and might still be operating with resources calibrated for previous levels of activity. Could these teams be overwhelmed by a sudden surge in demand, like a conveyor belt speeding up without a stop button, unable to sustain a positive user experience?
Innovation caught-up and passed you by
In the rush to work with new, shiny tech, teams can become absorbed in the complexities of connecting data, algorithms, skills, and APIs to create what may seem like magic to leaders and users. Meanwhile, the rapid pace of AI advancement continues, pushing out older methods. After six months of developing based on an earlier model, a team may find that a newer release already includes the features they’ve been working on, but with better functionality and a lower cost.
Which Species of Team are you?
The AI chatbot team, after months of development, has started showcasing their work at group meetings and ‘Show & Tell’ sessions. To their surprise, they discovered that another team in the organisation had also built an AI chatbot with similar functionality. However, the other team’s version was developed faster and featured capabilities the first team’s chatbot lacked. The second team utilised a different AI provider’s tools, learned new skills, and followed a separate path, creating distinct solutions that do not easily integrate.
What this means for your AI project
- Consider the entire service journey when building any solution to anticipate where bottlenecks might emerge, disrupting customer service.
- Choose your tools carefully in an unstable market; it’s costly to back an early entrant that may falter, like Blackberry did against the iPhone or Betamax against VHS.
- Platform changes come with high costs in technology, training, and skills. Standardise on a platform and use it consistently across your organisation.
AI can accelerate productivity, but not if you hit every obstacle along the way.