Using AI at a company shouldn’t just come out of nowhere. There is a time during a company’s evolution when AI is naturally needed. Each company starts out as having a data ecosystem. The more their data grows, the more they need Business Intelligence in that ecosystem to better record, manage and use data for planning. The more planning and analysis they need to do, the more analytics needed in the ecosystem. Better planning and better analysis requires better understanding of their data. This means they need more algorithms. The more sophisticated algorithms they need, the more likely they need AI in their ecosystem. The evolutionary path of the data ecosystem to a AI-BI-analytics ecosystem strongly mirrors the progression through a company’s capability maturity model.

Additional support is needed when applying AI solutions. What makes a solution AI is its ability to learn. AI uses datasets to learn. These datasets need to be structured, managed, and delivered by BI. The datasets also need to be able to include analytics, results from algorithms, and results from other AI. To optimize an AI and its learning can require changes to BI, datasets, analytics, algorithms, and other AI solutions. This means any changes to an AI can create a cascade of changes throughout the AI-BI-analytics ecosystem.

The evolution of the AI-BI-analytics ecosystem mirrors the progression through the company’s capability maturity model. This also means leadership and how leadership evolves follows suit. Often leadership priorities do not change. What changes is their top priority. When ecosystems have a BI emphasis, managing costs is the top priority. When ecosystems emphasis analytics, the top priority is managing opportunities, insights, and rankings. When ecosystems emphasize AI, the top priority is managing risks of the company and its ecosystem. AI makes mistakes and those mistakes expose the company to risks of unexpected costs, revenue losses, and litigation. The more companies delegate planning, decision-making, and manual work to AI, the bigger the impact AI mistakes will have on the company, its customers, and the market.

There is high demand for talent to support AI-BI-analytics ecosystems. To meet this demand, schools are preparing learners with a wide breadth of tech approaches. However, there are challenges when it comes to AI. There are time constraints for courses that limit how deep learners can go into the application, experimentation, and optimization of AI solutions. And unlike other tech, AI has significant advancements every few months. It takes additional investment to keep courses current and relevant. Schools recognize this and are providing courses and modules dedicated on how to agnostically apply AI and disruptive tech. The challenge now for these learners is when they apply what they know in the work environment. It is extremely easy to misapply AI. It’s easier still to recommend AI solutions that don’t fit the company in terms of its management style, culture, and influential personalities.

Remote work is a solid option for reducing costs to business infrastructure. However, not having people co-located limits various forms of knowledge transfer and skill development. Guardrails are needed to enforce workers to follow strict procedures for knowledge sharing, skill development, and learning Do-It-Yourself (DIY) approaches. Web searches, web content, and AI are not strong substitutes for traditional knowledge transfer. Anything that involves algorithms in any shape or form prioritize what’s popular over what’s correct and right. Further, without the right polices and procedures, remote work can lead to less productive teams, throwing work over the wall, and reinforcing weak bonds between colleagues. Practicing out-of-sight out-of-mind is extremely easy to do in a remote work structure.