Properties of a Fine AI Company
In the next decade, the standard for modern companies will be an AI company. This means that organizations must seek to utilize AI as part of their core offering on the client side as well as in their key internal operations. Companies that previously built big data assets can now actively pursue this line of development. This will inevitably lead to competitive advantages as the possibilities of AI technology quickly evolve and deliver benefits for customer experience, as well operational excellence.
Being an AI company seems to be a lucrative positioning for now and tomorrow, but what does it take to become an AI company, and how does one become an AI company?
Just using AI for the sake of AI is plain silly. In fact, it could utterly destroy your business. Without the right management model, all current AI solutions are not smart enough to save themselves, or anyone else.
Ask yourself, “what role will advanced technology such as AI play in your organization’s future?”
To justify their place in the value chain, AI-based solutions must prove their value. The AI-driven business must outperform earlier technology by performance indicators that are set by business, not only engineers. In other words, this operation must also be fundamentally data-driven.
What kind of an impact can you expect?
The impact of AI can come in from multiple angles. In operations, AI can drive major changes in logistics, production, platform interfaces, sales, and marketing. Within products and services, AI can have a major impact on how services are catered, the amount front and backstage human effort, as well as the essence of the service. Opportunities to utilize AI are constantly expanding. For those unfamiliar with AI applications as a whole, I have outlined in another article how recognition, prediction and creativity are the three basic functions that AI systems provide, particularly for customer-facing services.
It is understandable that people may be both scared and anxious to see AI in action in their organizations. Although several AI victories have been widely celebrated over the years, prominent people such as the late Stephen Hawking have issued warnings of super AI dominating humanity. When will AI totally transform your organization?
Don’t hold your breath, I say.
Although AI solutions do outperform humans in selected tasks, they are extremely narrow in scope . AlphaGo only plays Go, and it does not press play on its own. A previous success story, IBM Watson of Jeopardy! fame, also turned out to be a disappointment. IBM used years and billions of dollars to develop an AI platform around Watson as well as branding cognitive computing, with relatively meagre results in terms of business as well as applications.
AI success is not generalizable
The takeaway is that current AI success is not generalizable. Although we can justify beliefs that an AI solution which has learned to spot first and second class cucumbers could do that for pumpkins, it would be a different thing to apply it to pre-owned vehicles, for instance. But even pinpoint narrow AI can drive notable gains with smart aim.
Strategic AI company, tactical AI
Just a few years ago, many companies were wooed by Big Data and associated hype. AI hype is a natural follow-up to that. The recent success of AI is built upon utilization of big data assets in combination with increasingly more affordable computing capacity that enables technologies such as Machine Learning to succeed. In order to get data hungry AI solutions working properly, properly formatted and accessible data in large quantities are needed. There are exceptions, but properly organized data warehouses or data lakes are never wasted in the hands of AI professionals who develop AI applications. So data maturity should be the first thing to check off from your list for getting started with AI.
To get further with the benefits of AI, organizations must actively seek to employ it. Particularly for old organizations, this can mean major changes in core technologies. Thus the change must begin at the strategic level.
Operational excellence with AI depends on a number of strategic and tactical choices. At best, this starts with an AI strategy or any other strategy that can advance the organization’s capability to employ AI.
On the tactical level of an AI company, I would place the technology stack and infrastructure choices. Good infrastructure tactics include resiliency and flexibility. Extensive microservice architecture and event-driven, serverless implementations are some of the architectural tricks successful companies play.
When the right enablers are in place, a company can gain the desired benefits from AI on the operational level. It is, of course, possible to start implementing AI in operations even without a good strategic alignment, but this usually means more (redundant) work with Data Engineering and similar enablers at the function where AI is being developed. In practice, the desirable characteristics of AI implementations are not unlike for any other software:
- Test everything exhaustively
- Get stuff out quickly for market validation
- Align well with existing, earlier technology generation operations
- Scalable, robust and fault tolerate
Who wants to be an AI company?
Are companies truly seeking to position themselves in this way? International tech giants and former internet companies are, naturally. Those who hold a massive amount of customer data are the best positioned to use it effectively.
In Finland, the leading financial group OP is seeking a competitive edge through data. Their strategy is centered on customer experience which they want to support with the utilization of customer data. According to group President Timo Ritakallio, they believe customer data will become the key asset that helps them to build customer-facing, superior services as well strengthen the operational pipeline of internal products. AI is a key mechanism to help to create value out of that data and to create services that are empathetic towards customers.
If you are willing to raise your hand and jump on the moving train, act now. Tomorrow’s AI company is born today. The performance and maturity of AI solutions has recently taken huge steps forward. Although their application is still narrow, they can already provide quick returns on investment and thus secure first mover advantage for those who are willing start their AI journey today. When you begin, start by looking at the items we mentioned: data strategy, AI strategy, AI tactics and operations, Data Scientists. Which one’s do you have, and which are you uncertain of? As a consultancy, Fourkind naturally sees that external partners such as us are perfectly suited to help you start the journey.
Fourkind is the hands-on advisory that excels in strategic thinking and implementation of business-critical emerging technology applications. We also help guiding our customers to grow on the path to data-driven companies.
Thanks to Bruce Ferguson and Atte Jääskeläinen of Fourind for editing!
Author Lassi A Liikkanen works at Fourkind as a designer, dedicated to improving customer experience with data. Twitter:@lassial
