Start, Learn, Grow, Repeat: AI Success Simplified

"All conversations start with AI and most of them go sideways!"

This has been my observation in extensive interactions with businesses and leaders around the globe. While there is an acute awareness of AI's potential, there is also widespread confusion around a simple question:

"Where do we start?"

If you are feeling overwhelmed at the start of your AI journey, you are not alone. Amidst a sea of AI possibilities, organizations often paralyze themselves with the belief that they must overhaul their culture, upskill their entire team, and purify their data sets before they can effectively utilize AI. This mindset of waiting for a perfect scenario only leads to lagging behind in a rapidly advancing AI landscape.

"So, what should I do?"

In matters of AI, no one has all the answers. I can offer what I believe are essential considerations, highlight pitfalls to steer clear of, and present a straightforward yet efficient framework to kickstart your journey.

The most critical pitfall to steer clear of:

Do not start with the daunting task of becoming an AI-first or data-first company!

I strongly encourage companies to adopt a more realistic and achievable approach when it comes to integrating AI into their operations. The reality is that AI technology is advancing at an unbelievably fast rate, one that far exceeds the pace at which most company cultures are able to adapt and evolve.

This significant pace discrepancy between AI advancement and organizational adaptability often renders traditional integration strategies obsolete and ineffective. To bridge this gap, businesses need to focus more on incremental and manageable steps towards AI adoption that align with their current cultural and operational capacities.

Avoid the low hanging fruit!

The allure of low-hanging fruit is undeniable. We've all been there—craving a quick win. However, resist the temptation!

In the realm of AI adoption, the goal is to attain substantial value. If the use case doesn't matter to you, applying AI to it won't make it any more significant. Choose a use case that holds deep importance for you and your organization.

At the same time, you don't want to try to spend too much time overhauling organizational culture, processes and data sets before you tackle the technology. While some modifications to your current practices will certainly be necessary for integration and optimization, a total revamp is not required. In this instance, you have to match technology to your culture and needs and not the other way round!

The sensationalism surrounding AI in the media is also distracting and can potentially leave you hung up about its possible implications and impact. While it is true that AI will change everything in the future, there is value in the present that you don't want to miss.

"So, how do I select the use cases?"

Instead of getting bogged down by grandiose plans or future predictions, start with a couple of specific use cases.?I recommend adopting an 'AI-iterative' mindset and dividing organizational functions into three categories:

1. Functions to Accelerate or Augment: Look at what your organization does everyday and identify tasks that could benefit from acceleration. These accelerations will definitively enhance efficiency, saving time, money, or improve employee well-being. Make sure that the answer to the questions "Did AI help in this task?" is a binary yes or no.
2. Functions to Replace or Avoid: Determine which tasks are so mundane that it's best to have an AI handle them. An example is generating a first draft of a policy for a new security tool. Again, the value is a straight forward efficiency.
3. Functions to be Created: These are novel tasks your organization has never undertaken before – the things AI will enable. However, resist the urge to start with these tasks first as it can lead to confusion and misdirected focus.

Once you have a shortlist, evaluate feasibility by reviewing each selected task against its data set requirements. Aim to pinpoint high-value functions that interact with minimal change management and data requirements so you can deliver maximum value through use cases that are practical and feasible to implement.

Select ten use cases in the first two categories, analyze some more and down select to 5 and then pick one. Yes, one!

The benefit of this segmented approach is that your initial AI use cases are well-defined with tangible success criteria. Small but definitive victories will increase your confidence and understanding of AI, build trust in the technology, and generate value from day one without disrupting current operations. This will also help with mindshare gains within your organization(s); justifying further investment, and keeping learning and value creation continuous.

Sound familiar? It is!

My regular readers will instantly recall this as yet another application of the ARC framework(Read more about it here). The ARC framework acts as a blueprint for understanding technology trends philosophically. Here, it's a practical application of the ARC Theory, serving as a useful guide for identifying suitable AI use cases.

While AI itself is a revolutionary technology, its adoption is going to be evolutionary. None of us ran a marathon the first day we picked up running!

Embrace the opportunity to join the ranks of those who are leveraging AI to shape their futures.

But start within your current capabilities!