Businesses are adopting AI backwards.
They start with the technology.
Which AI should we buy?
Should we build an agent?
Can ChatGPT do this?
Should every employee have an AI assistant?
The better starting point is not AI.
It is friction.
Look at your company and identify repetitive decisions, repetitive communication and repetitive movement of information.
Where does somebody receive information in one place and manually enter it somewhere else?
Where are employees answering the same questions every week?
Where does someone spend hours preparing a report that management reads for five minutes?
Where are leads forgotten because a person has to remember to follow up?
Where does information exist, but nobody can retrieve it quickly enough to make it useful?
Those are usually better AI opportunities than trying to invent an impressive chatbot.
A simple framework is to divide work into three categories.
- Judgment work requires experience, context and accountability.
- Process work follows relatively predictable steps.
- Information work involves finding, organizing, transforming or communicating information.
AI becomes extremely powerful when it removes unnecessary process and information work so people can spend more time on judgment.
But there is another side companies underestimate.
Automating a bad process does not make it good.
It makes the bad process happen faster.
Before implementing AI, ask:
If we removed the human effort from this process tomorrow, would we actually want the process to remain exactly as it is?
If the answer is no, redesign it first.
The companies that benefit most from AI will not necessarily be those with the most tools.
They will be the companies that understand their operations well enough to know what should be automated, what should be redesigned, and what should always remain human.
Technology is leverage.
But leverage only becomes valuable when applied to the right system.
Author: David Castañeda Jaquez