Enterprise Artificial Intelligence: 10 Commandments for 2018


If there is one technology topic that drew a great amount of media attention this year, it is artificial intelligence (AI). Thus far, you can say that only digital technology giants such as Amazon, Facebook, and Google—along with a few other early adopters—have tasted success in implementing AI technologies. Here are 10 commandments to finally get you started on your enterprise AI strategy.

  1. Get off the sidelines. If you are a typical enterprise, so far, you have been watching AI from the sidelines. But AI technologies are maturing and are now available for commercial deployment. You should start to craft your enterprise AI strategy sooner than later.
  2. Be realistic in your assessment of AI software capabilities. Enterprise AI applications are making steady progress, but, unfortunately, the hype has gotten ahead of it. Spurred by relentless media coverage, expectations are unreasonably high. Evaluate the technology skeptically, and understand that value comes from data, not just software.
  3. Rise above FOMO. Don’t let a vendor or a consultant use the “fear of missing out” to goad you into purchasing AI technology that you neither need nor can effectively exploit. The temptation to rush into a project you have not clearly thought through is heightened when your C-suite colleagues take an active interest in AI. If you are a CTO/CIO, avoid AI initiatives just so that your team can be seen to be doing something in this hot area.
  4. Start with opportunity assessment. Assess how AI tools can improve your customer experience, operations, and productivity in the context of your business goals. Identify opportunities to integrate AI capabilities into your business processes. Revisit the opportunity identification every few quarters as it can undergo rapid changes.
  5. Don’t underestimate the complexity and effort. For the typical organization, it is still difficult to implement and realize the benefits of AI. Pilot projects may work well, but to cross the chasm from proof of concept to production remains elusive for many because of technical complexity and resource intensity. Prioritize and focus on a few key areas rather than trying to start several projects all at once.
  6. Fix the fundamentals first. There are several ingredients needed for success. For example, AI relies on availability of good quality data, but enterprise data is locked up in silos. Fix any data integration problems. Data cleansing and preparation are musts for any AI project to succeed. Include this in your project plan and budgets. Also, do you have a scalable data-processing infrastructure?
  7. Think incremental improvements, not wholesale disruption. Current AI technologies are in their infancy when it comes to higher-order cognitive capabilities. But they’re reasonably good at narrowly focused tasks. These activities may not necessarily constitute end-to-end business processes but can still provide enough of a productivity uplift to justify your investments in AI projects. Narrowly scoped projects will likely result in more successful business outcomes.
  8. Distinguish between automation and AI. There are automation tools for a variety of standardized and easily codified work tasks. Hoping to capitalize on the interest in AI, several vendors are opportunistically rebranding themselves as vendors of AI tools. Such process automation tools have their uses, but they are not AI tools, and they need a different implementation approach.
  9. Get your core teams future-ready. Not everyone on your technology team needs to be an AI guru, but form a core group with expert resources to lead the AI charge. Include business, technology, and change management experts in this group, since AI projects are not technology projects, but they involve cross-functional collaboration.
  10. Blaze your own trail. When it comes to enterprise AI, the learning curve is not only steep, but unique to your organization. Try different approaches with an understanding that even not-so-successful projects enrich the enterprise knowledge corpus and smooth the path for your next set of initiatives.

Related Articles

Marketers can be excused if they are facing cognitive dissonance about artificial intelligence (AI). They may be struggling to implement foundational marketing/technology tools, but they are expected to be ready for new AI applications that are way more complex.
It's 2017, and everyone is wrestling with fake news. Given the wide-ranging nature of artificial intelligence (AI) use cases, it is only natural to ask whether machine learning can help us combat fake news. It should be fairly simple to build ourselves a fact-checker, right? After all, software is able to flag spam email or pornographic images with a high degree of accuracy. It turns out that machine learning techniques are useful, but they have their limitations—and we are far away from a fully automated fact-checking tool.