THREAD: Go Strategy First in Your #AI Projects, or Pay the Price

Fun fact:

If your #AI project is predicated on solving one discrete problem (i.e. a band-aid), you're inordinately likely to fail.


These kinds of promises come back to ruin AI initiatives, and create rifts in vendor/service provider relationships:

-- Promises to deliver X measurable result
-- Promises to complete the project within X time horizon
-- Promises that X capability is definitely build-able

These promises reinforce the WRONG message to executives and stakeholders, they reinforce the idea that "AI is IT"

...when in fact...

"AI is NOT IT."

@calccon does a great job of explaining this point in this @Emerj podcast interview: https://soundcloud.com/emerjairesearch/pitfalls-to-avoid-for-the-roi

Pandering to executive AI ignorance might win a pilot, but will hurt relationships and is unlikely to lead to a deployment.

It hurts the vendor or service provider (no lucrative long-term partnership) and the buyer (no real lessons learned or strategic progress).

An AI initiative needs to tie to a strategic mid or long-term objective, and needs to be seen as an investment in learning (new ways for teams to operate, new ways to handle data infra, etc. https://emerj.com/ai-executive-guides/critical-capabilities/). Almost everything else is disingenuous.
You can follow @danfaggella.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: