TrueState is a technology company focused on empowering organisations to solve their most important problems with machine learning and artificial intelligence.
Our developer platform provides teams with access to cutting-edge algorithms, high-leverage APIs and an easy-to-use interface, while our implementation solutions help organisations take their first steps with AI with confidence and peace of mind.
“I’m worried AI will take my job” - 22% of American workers according to a recent Gallup poll.
This fear, known as automation anxiety, isn’t new. We’ve seen it from the luddites in the early 1800’s fearing the rise of machinery in the textiles industry, to advancements in manufacturing in the 1930’s, to the rise of the internet in the late 1990s.
While there is widespread consensus that AI will create more jobs than it replaces, that doesn’t change the fact that automation anxiety is real, and that it’s a killer for AI transformations. I’ve seen it first-hand.
Having worked through countless AI transformations at McKinsey & Company and in Private Equity across a range of industries from Tech to Consumer Finance, Pharmaceuticals to Telecos, trust me when I say that overcoming automation anxiet is of the main hurdles organisations need to overcome to successfully implement AI use cases.
I’m not naive to the fact that the shape of the workforce will change as a result of the evolution of AI technologies. Specifically, I think that the days of giant customer service functions are nearing their end, with businesses running these large teams undoubtedly shifting to an AI-dominated model over the next 10 years.
For most companies, the highest impact opportunities from AI come not from replacing employees but from empowering them to make better decisions - which sits a lot better with workers and bodes well for executives trying to reap the benefits of the technology.
There are myriad potential use cases for generative AI and classical machine learning within organisations; however, selecting use cases that empower existing employees in their current roles can be a game-changer. This approach not only boosts productivity and mitigates transformation risks but also ensures long-term value creation.
It’s the job of management to identify the opportunities that generate the highest return on AI spend, that also have the highest likelihood of being implemented.
This process need not be ambiguous — at TrueState we follow a standard process to identify these dream use cases and successfully roll them out. Here’s what it looks like (in a very simplified form):
Step 1: Identify Key Processes
Begin by pinpointing crucial employee processes that impact the bottom line, such as sales, account management, and supply chain management. If you’re stuck, start with your cashflow and income statements.
Key point: Once you’ve identified the processes, check that employee incentives are aligned with business success. If not, you should fix this as it will ensure both you and your team will be gunning for the higher performance that AI can deliver. To take from Charlie Munger - “show me the incentive and I’ll show you the outcome”.
Step 2: Engage with your team to identify the key decisions
Go beyond processes and identify the critical judgement calls that drive success. Whether it's selecting the best message for a new lead, engaging with clients to prevent churn, or timing stock orders to avoid shortages, understanding these decisions is crucial. Explore the frequency of decisions and also the volume of data surrounding them.
Why? Because it helps uncover the best model types to use.
Take a sales processes for example. If your biggest challenge is identifying which leads to talk to, a classification model often most appropriate. If you’re identifying the perfect topics to engage in you might be working with generative models, classifiers and optimisers all in unison. If you’re trying to engage with self-sign ups and close leads without human intervention, more complex generative flows might be more appropriate.
It’s not important that you remember the types of models used, but rather that you’re explicit in the frequency, format and supporting data for your most important decisions.
At this stage you should be able to also estimate the likelihood and impact of marginally improved decision making, which enables dollar-for-dollar comparison of initiatives and more robust prioritisation.
Step 3: Choose the right algorithms
Selecting the appropriate AI algorithms is often more art than science. A reliable solution partner should help you identify a robust solution and deliver initial results quickly. If you're waiting more than a month for your first outcomes, reconsider your technology partner to avoid deployment challenges.
Step 4: Build with Employees, Not for Them
Involve stakeholders early in the process and design systems based on their input. Regular development sessions and active listening will make the system more familiar and reduce the "alien effect" that AI can sometimes create.
Step 5: Find a Champion and Release Slowly
Identify a passionate solution champion to advocate for the technology. Ideally you should look for a respected member of your operational staff. When you’re rolling out the technology, take your time, as the initial results and format might not be perfectt and you’re likely going to need go iterate on the solution with your champion. As you’re getting results in communicate performance improvements clearly, frequently, and loudly. A gradual rollout, accompanied by continuous feedback, ensures smooth adoption.
Step 6: Continual Feedback Loop
Post-implementation, keep listening to your team. Are they enjoying the new process? Are they meeting their targets? Could the recommendation format be improved? Ongoing feedback is vital for continuous improvement.
Conclusion
AI has the potential to significantly elevate careers and break down traditional barriers, but its success hinges on effective change management and alignment between management and employees. By following these key steps, you can better ensure a smooth AI transformation that empowers your team to achieve new heights of performance.
Photo by Matt Artz on Unsplash