Developing an AI strategy, the human+ approach - how to get started
The rise of the machines has long been prophesied – even before Schwarzenegger had his say in The Terminator franchise. Today, with the synchronised upsurge in maturity and accessibility of the model / compute / data trifecta, we seem closer than ever to a time where machines will replace humans and run the world for us.
Are we truly that close though?
Not really. Unless you’re Arnie of course.
Replicating human-level intelligence that matches our ability to understand the world, to reason, and to make complex decisions, is still a distant dream. Equally, AI lacks an intrinsic motivation beyond the one we code for it. Without that self-awareness and intention, we still need the human to be (mostly) in the loop for many more years to come.
However, one thing is clear: humans with machines will replace humans without machines. Indeed, they already are.
When it comes to the business-level application of AI, especially in the small to medium enterprise, there is so much untapped value available to support business growth. The question that often stops an organisation from developing their AI strategy at pace is how to get started.
This article sets out the Human+ approach that SMEs can begin to adopt now, the things to consider in choosing when and where to begin as well as some guidance on that all important question of getting started.
A Human+ approach is the first mindset to adopt when considering how to use AI in your business today. With the traditional and generative AI capabilities currently (and widely) available, human plus machine opens up a lot of possibility. This approach also plays right to the heart of adoption – removing the fear of jobs being lost and replacing it with a more positive motivator: what can I do to make my work more efficient every day?
Again, with a Human+ approach it’s essential to shift your thinking from AI as a technology problem when it’s actually one of business integration. Most organisations are not on the front line of AI development, and they don’t need to be. Given the rapid pace of change in Data and AI systems right now, deliberately choosing to sit in the early majority category on the diffusion of innovations curve is a perfect place to be. Forget what’s coming, how do we start using what’s there?
The final element to be very aware of is time. Like any good strategy, it’s so important to take the measure of time. Just because the AI Revolution has started, it doesn’t mean that it needs to start for your business right now. Consider what else you have going on – perhaps a PostgreSQL database migration that’s consuming technical resource or an investment round that really needs to demonstrate you’re on top of the AI agenda. The question you need to ask is should your company be betting on AI as a growth strategy today?
The agility of a smaller organisation makes the perfect growing ground for a Human + approach. Whilst the definition is fairly broad[1], there are 5.51m UK organisations classified as SME, accounting for 99.9% of UK private sector businesses, of which the vast majority - 99.3% - are classified as small (0-49 employees).
When it comes to developing your AI value journey, the smaller organisation offers many advantages:
1. There’s an openness to change and new ideas. With fewer scars from previously failed projects - digital transformations, cloud migrations, next generation software implementations - the traditional change fatigue of the big organisation doesn’t exist, and the opportunity excitement is real.
2. Teams know how sh!t gets done. There’s nowhere to hide, they know the workflows and processes that truly add value to their customers and, therefore, they know the areas where AI can have the greatest impact. This avoids the burnout and paralysis from endless POCs that go nowhere in department after department of the large corporate.
3. Leaders can act quickly and decisively. With a flat organisation structure and easy access to the board (if one even exists), business leaders can make the decisions on strategic investments, priorities and resources. They don’t need a Chief AI Officer or AI Tsar to centralise control and dictate the agenda. They can quickly get new initiatives into the hands of frontline teams all by themselves.
4. There’s little need to “go live”. Many SMEs are cloud and data natives, they know how to make and access “tools” or “hacks” that add value to a business process without the need to make it “enterprise ready”. Prototypes of intelligent agents can be readily picked up and used in a real live business process without all the test to production rigour that exists in larger entities.
Taking the right step in the right direction begins with having a plan that fits your organisation today. As next generation leaders, begin by asking some of the Human+ questions posed above. A quick note: the approach outlined here is not a one-size-fits-all. It’s tailored towards a Human+ model for small to medium enterprises that firmly believe AI will play a part in the future of business growth for them. Let’s get into it.
1. Level up your understanding. Everyone is in pathfinder mode right now, trying to work out where to go and how to get there. Time to evaluate the environment and get a base of knowledge that can help you navigate. Read, consult, ask stupid questions, seek out education, court suppliers. Whatever you can do that works for you. You’re building up an understanding of the different types of AI solutions and systems; how they work; where they can be applied; the patterns of problems they solve; what they need in order to be successful. For example, task level productivity tools such as Co-Pilot are very different to process automation tools with some embedded AI like Blue Prism. Without an appropriate level of education and awareness, you can’t adequately gauge your risk appetite towards AI and, more importantly, you will struggle to define what AI means for your business. Don’t forget that everyone should be behind this - don’t just delegate to one person or one team, get the whole business community behind the research.
2. Find a valuable problem to solve. Candidate tasks, issues, projects and growth priorities may vary, but the goal should remain the same: identify AI’s potential impact on select business processes then break it down into measurable and improvable outcomes that have value and can be tracked. Start in a narrow way then widen out the thinking to create a backlog of use cases that will become the outline of your future roadmap. Above all else, avoid the common mistake of making AI a technology solution waiting for a problem to be solved.
3. Lay the ground carefully. Understanding the full ecosystem that needs to be in place is very important. Processes and governance, technology platforms and tools, security and compliance, people and skills, internal and external data. That doesn’t mean you need the whole thing in place before you begin, in fact, you absolutely shouldn’t. You just need to be prepared. The AI landscape is changing so fast that by the time you get it all done, the world will have moved on again. Instead think of building out iteratively. Get the basics in first - the components that allow you to get started - and have a view on the next set of components required to grow further.
4. Run experiments that matter. Now is the time to make the conscious and informed choices on which experiments to run first. Start small but think big – How Might We…change the way we start new customer conversations with increased personalisation in our welcome emails? Smaller use cases that succeed will build confidence in the process. Smaller use cases that fail bring new learning on how to get it right next time. Don’t be afraid to “hack” solutions together to get them into the hands of the people who will be using them but just make sure that you set realistic goals to avoid the risk of losing trust and stakeholders.
5. Anticipate the future risks. But don’t go too far! Being bold and trying to shape the future of business growth is not without risk – but you should be used to that now as entrepreneurs in your own rights. However, AI does bring its own set of unique risks that need to be anticipated and managed. These risks increase when you’re building intelligent agents that are customer-facing and handle customer data. Some of the main things to be aware of at this stage include data privacy, bias, security, explainability, robustness, hallucination and issues with logical reasoning. Told you not to go too far!
6. Build talent in-house. Absolutely use a third party to give you a boost and get the ball rolling on capability build-out but never, ever create a dependency on them. You must invest in development and training programmes for existing staff or hire fresh talent into your current teams. Only those that truly understand and believe in your purpose and your mission will support the culture of growth, experimentation and commitment needed for success.
7. Scale-up together. Once you’ve proven the value of focusing your efforts and resources on integration and process change, it will begin to sharpen what you look for in an AI system at scale. You truly begin to understand the importance of seeing your data and the design of your Data & AI architecture as truly differentiated and competitive assets. Furthermore, you end up leading the whole organisation to drive more and more experimentation or testing that in turn will feed the overall intelligence of your AI system at scale.
8. “Culture eats strategy for breakfast”. The greatest line Peter Drucker never wrote! What he did say though is that “culture – no matter how defined – is singularly persistent”. At the end of the day, no matter how good your AI approach is, it’s the people who implement the plan that matter most. Get the whole company behind the push to the future and develop the right talent with an element of competition in their experiments (who doesn’t want to be the best). The most important way to build your AI programme is through your own culture: it’s adoption that will constantly activate and inform the intelligence behind your future AI systems.
In the words of the Greek goddess and innovative running shoes: Just Do It.
The only failure is the failure to start. This truly is the Age of AI. How and when you choose to develop or use intelligent agents and AI systems depends entirely on your business and the growth goals you’re targeting.
But make no mistake. This changes everything. Moving fast isn’t a luxury or even an option, it’s a necessity for all businesses. Doing nothing means you’ll either lose a major cost advantage or miss out on a growth opportunity to another organisation.
Time to equip the humans, the machines really are rising up.
[1] The UK government definition of SMEs encompasses micro (less than 10 employees and an annual turnover under £2 million), small (less than 50 employees and an annual turnover under £10 million) and medium-sized (less than 250 employees and an annual turnover under £50 million) businesses.