Written by: Nino Ilveskero, Talent Base Oy
Every growth-hungry business will soon have a schedule for being AI-ready – for having all the skills and tools to implement cross-cutting and company-wide AI systems. In this post, I will explain in plain words what we at Talent Base mean by AI readiness.
We believe that a company or organization is AI-ready when its strategy, capabilities, and processes support large-scale use of AI. By large-scale, we refer to applying AI in ways that have a notable effect on the organization’s performance.
Make sure you have a clear strategy
Implementing AI is always a big leap for an organization. It requires tremendous courage and major investments. The organization needs to adjust its processes, acquire the talent and upgrade the technology. All these investments are necessary before being able to reap the benefits of AI.
It would be easier to stay put and see how it goes for others. However, the ones to utilize AI first will have a better start in the race, so hanging back while others charge forward may turn out to be fatal. To put it simply, doing nothing is the foolhardiest thing to do, as you would be staking the future of your organization on the failure of your competition.
Manage the data, leverage the technology and train the people
The most important elements in the use of AI are data, technology and people. Without proper Data Management and Governance, the I in AI doesn’t stand for intelligence but an idiot. Without the people who set the goals, train and supervise the system, the algorithms do not create anything new and meaningful, but the system becomes random and useless.
The biggest benefits of AI systems as decision-makers are speed and flexibility. An AI-ready organization with good governance mechanisms can quickly adjust and reshape its AI systems.
Recently, Vuokko Aro’s tweet about Google Translate’s gender bias went viral and stirred up another discussion on the challenges of machine translation. This is a good example of why functioning AI governance is needed. In our opinion, an AI-ready organization can quickly steer and teach AI new types of decision-making models.
From data governance to AI governance – three lessons learned
Talent Base has been working with data governance for over ten years. In AIGA, we are applying that knowledge specifically to AI systems. Although AI governance has its own characteristics, many of the lessons apply.
Automate high-quality inputs
Having a data governance model ensures that the AI developers always have access to the curated and actively maintained datasets and don’t need to waste time monitoring data quality. Right now, the classic data governance model has become an important and so far the smartest way to manage data assets. It helps organizations keep their data constantly usable for AI applications.
Embrace life-cycle approach instead of managing single applications
Earlier, data was mainly gathered for a specific purpose and changed only moderately. As the use of AI keeps increasing, data is created, used, and reworked a lot more. The data life cycles are becoming more complex with multiple twists and turns, posing new challenges for quality control. User- and service-oriented and continuously evolving data governance model also helps to manage the data in an agile way.
Eliminate complicated subprocesses
The data governance model may seem laborious at times, so it might seem tempting to divide tasks into several siloed subprocesses. However, this leads to a vicious circle of fixes: whenever a problem in data management is tackled, a new problem will pop up somewhere else. When decisions are made by people only, this is not such a big problem, as people can see errors in the data and find ways to patch them up.
When AI is taught, it is given data that guides it towards certain types of decisions. When the data is in good condition, the organization can reteach its AI at a moment’s notice and so enable it to make better decisions and to react to changes as quickly as possible. In most lines of business, this means gloomy times for competitors.
The writer is a partner in Talent Base Oy as well as the AIGA steering group member. He has worked for over 20 years with digital services and IT system development in different industries.