Why you need to think about AI (even if you’re already sick of it)
AI is everywhere. It’s impossible to read or watch much without seeing an article or advertisement related to AI. Because of this ubiquity, we at Chartis recommend all our clients have an AI working group and policy. Regardless of any active investment in AI by your brand or company, someone, somewhere in your current process or workflow is using AI. Even if it is just using ChatGPT to summarize meeting notes it’s important to acknowledge and define any desired guardrails around use and transparency.
When creating a working group and policy, we recommend taking a cross-functional approach. While it might be tempting to leave it up to the Legal team, we recommend participation from IT, Marketing, Product and Customer Experience. This wider set of perspectives will ensure a more holistic consideration of any intended and unintended ripple effects of AI use. An additional benefit of creating a working group is having them primed and ready to go when considering direct investment in AI.
Where to start with investment in AI
At Chartis we recommend taking a Use Case First approach to AI – start with the need, and then define the solution, not the other way around. Think first about good problems to solve for your customers and/or business, and then consider what AI-powered solutions might solve those problems.
At present we see three macro AI use cases… things AI does well:
- Identification - the tool can recognize and surface content and data based on their similarity to a defined set of data.
- Creation - the tool can visualize and describe (in words or pictures) new content and data based on its similarity to a defined set of data.
- Decision Making - the tool can choose optimal content and data use based on the likelihood of achieving a defined set of goals.
You don’t need to start with a solution that covers all three. We recommend starting with your customer relationship to establish use cases of value. If your customer relationship is built on convenience, would making more accurate, timely decisions improve the relationship? If your customer relationship is built on security and trust, would faster threat identification improve the relationship? Understanding what delivers value to your customers is essential to delivering value to your bottom line through investment in AI.
How to test, learn and scale with AI
Once you’ve identified your use cases, it’s time to define your solutions. At Chartis we start with three key questions to define AI solution spaces for our clients.
- What data do we need to fuel a solution?
Answering this question can assess the cost of the solution space – is our required source data something we already know, or something we need to invest in to find out? - How stable and accurate is the data?
Answering this question assesses the feasibility of the solution space, while AI models continue to evolve, most of the currently mature solutions rely on prediction in some form - data accuracy and stability can reduce risk in output. - How much risk is acceptable?
Answering this question is a key decision point in the viability of a solution space – if the AI solution fails, or generates unintended negative consequences, how much exposure does it create?
Answering the above questions will also inform a build or buy decision. We work with AI vendors across a wide variety of solution spaces. Beginning with a third-party tool can reduce risk and allow you to start small. Starting small, to address a well defined use case, allows you to learn and adjust ahead of larger roll-outs in scale, scope or both.
Above all be clear
Finally, a note on transparency – at Chartis we believe that transparency in AI use is fundamental to success. Trust is currently the biggest barrier to AI adoption with 67% of respondents reporting low to moderate trust in AI in a recent KPMG survey. Be transparent both with customers and internal stakeholders whenever you employ an AI solution. Transparency brings us back to where we started – AI is everywhere and likely already in your workflows, active investment or no. Wherever you are starting from, at Chartis we’d like to help.