Making waves, not ripples: Scaling AI initiatives that redefine companies – Part 1

During the past decade, AI has become front and center in the minds of many leadership teams, and Silo AI has been broadly sought out to discuss how to leverage AI for value creation. From these hundreds of discussions with business and technology executives over the years, Silo AI CTO Niko Vuokko has now distilled some of the key decision points and learnings in the form of a multi-part blog series, diving into the profound challenges businesses face in adopting and scaling AI.

Beyond the checklist, the dilemma businesses have with AI

We keep on hearing how AI is a big thing. Even then, despite sounding naive, we often feel obligated to tell our clients that it will be a far bigger thing than most expect. And while it hardly feels like the AI progress and hype could reach any higher pitch, the obvious fact remains that we're only in the very early stages of what's to come.

So, what can we already achieve today? And what do we even mean by AI? With the scale and breadth of Silo AI's experience, our past work is a rather good example of where AI is already getting into real-world use. Other than the broadly talked about AI use cases in autonomous/assisted driving, generative-AI-driven healthcare advisory, or visual quality inspection of COVID-19 vaccination vials, we've covered a wide range of AI use cases in chemical sensor calibration, financial derivative analysis, pharma molecule assessment, supply chain optimization, mining drill control, chemical process optimization, etc. This is a long and varied list of already useful AI use cases. So, how can we claim to remain in the early stages of AI deployment?

In many ways, AI today is similar to how things were 100 years back when the very first commercial airplanes came about. They were pretty creaky things, held 2 to 3 passengers, and could usually take you through the air to your destination. It was a bit risky, mysterious, and unfamiliar, but the feeling was truly magical when it worked. Building and using AI is often similar to this nowadays. It's not evident if and how it'll work, and it takes plenty of effort to get things going, but when AI works, it's again nothing short of magical!

Aviation has changed a fair deal since those days a hundred years ago. But it hasn't happened without relentless learning, massive investments, and significant changes in how the world operates. Similarly, with AI, we are now entering a multi-decade era of enormous industrialization, allowing us to integrate AI into our lives much like we did with electricity and the internet.

Let's take some extra rounds with the time machine, first further back when electrification was the front page poster child. So what was the proper response of a business to this new technology, and what happened? Was it rapidly installing electric light bulbs in the office, leaving the factory side as such, and calling the investment program a success? With the internet upheaval in the 1990s, companies quickly set up an e-mail address and a website listing their address and phone number. But looking back to either of these examples, it would be laughable to claim that focusing just on such actions was anywhere close to making the most of the technology or even the bare minimum of what was soon needed, not to consider what was implemented by the leading companies of the following decades.

Becoming more intelligent and more efficient isn't enough to save a business 

Significant changes in the world always open the door for new market entrants, and often, this happens by either breaking down or building on top of old value chains. Essentially, it does not attack only the technical basis of incumbent companies but their core DNA of how things are run. This certainly is the case with AI. If it were only about technology, the big enterprises of today would have overpowering advantages over anyone trying to enter their market with a "now with AI" pitch. 

Machine learning inherently depends on the underlying data, making the whole field hold very high economies of scale and giving existing companies with big data holds and even bigger balance sheets a good leg up against any new entrants. So it's not the technology that is the challenge, but how AI will change how markets work, customers behave, and what the company mission should even mean going forward. 

As picked up from the words of so many telecom operator executives, "we don't want to be just a dumb pipe" will be a very relevant reflection for many companies struggling to adapt to the new AI world. AI has brought an explosion in what we can expect from machines, but it has also brought a wave of investment and innovation. For many markets, we are already near the turning point where companies must either choose to become dumb pipes or find it in them to evolve. As much as installing light bulbs and websites wasn't enough to justify a company's ongoing existence in the past revolutions, giving your employees and customers access to more intelligent analytics and ChatGPT isn't going to help a company avoid the AI steamroller.

Many company boards know this already. The concern for the future exists, but the diagnosis and its treatment remain elusive. It's easy to say, "We need to act," but it's hard to identify the right plan and execution. Not precisely making this any more accessible, but many companies already have fatigue from digitalization efforts and burn marks from their initial attempts with data and AI investments that have produced lots of activity and spending, but less so in tangible bottom line.

Thus, a dilemma exists. How can companies devise a strategy on AI that can be 1) executed and 2) sold to its stakeholders not as yet another technology money pit but as a realistic and attractive business opportunity to pursue? For a proper cliffhanger, we'll leave a closer analysis of this question to the next part of our series. Stay tuned!

Stay up to date with the latest developments in AI

Silo AI helps the world's leading companies build competitive advantage with AI. As a blog series for the general corporate audience, many essential industry-specific (not to mention company-specific) analyses and plans of action will not be found here. Want to continue the discussion and get a real leg up on the competition for the long-term with AI? Get in touch, and we will help you get there.

Nordic State of AI report, our fourth annual report where we interview business and technology executives of industry-leading Nordic multinational companies, will be published in early 2024. The report offers business leaders, academics, and policymakers a comprehensive look at what's going on in Nordic AI. Subscribe to our newsletter to be the first to hear.