When people ask me why so many enterprise #AI projects fail, I am often able to sum it up in a single word:
"Expectations."
emerj.com/ai-executive-g…
Leadership isn't aware of what AI can do, what it's strategic value is. Worse still, leadership sees AI as just a new kind of IT. "Plug it in" and we're all set.
There are a number of reasons why AI is not like IT, here's a few that more boardrooms should understand:
** Time to Deliverable
IT: Hard to predict with new projects. Relatively easy to predict with repeat or well-understood projects.
AI: Nearly impossible to predict with new projects. Still very challenging to predict with repeat or well-understood projects.
** Results We Can Expect
IT: Relatively predictable, software can be very if-then, hard-coded - it will likely do what we program it to do, and enable the features and capabilities we expect.
#AI: Very hard to tell, experimentation required.
emerj.com/ai-executive-g…
** Near-Term ROI
IT: The functionality that the software enables.
AI: The learnings and critical capabilities that the team develops, such as knowledge of data infrastructure, building and functioning of cross-functional teams (data scientists and subject matter experts).
Consultants and business leaders who educate the C-suite on these topics (emerj.com/report/getting…) will have an edge in being able to not only lead #AI projects within the enterprise, but to win the resources and buy-in required to see them through to tangible success.