AI Management: People, Not Rocket Science

AI Management: No Rocket Science Required

• The constraint: A neuroscientist noted that the human brain operates with only about 20 megabytes of RAM—enough to process just four photos on a smartphone.

• The decline: NASA research suggests that by age 25, humans retain only 5% of the creative thinking capacity they had at age five.

• The lesson: AI should be used to simplify, not complicate. Use tools like CenturyLink did: they customized an AI tool to tell salespeople exactly who to call and what to say, turning a complex decision into a simple, successful action that increased sales by 10%.

• The mistake: Leaders often start by saying, "We need an AI strategy".

• The Harvey way: Figure out your business strategy before you invest in anything.

• Example: Li & Fung set concrete goals for speed and innovation first. By then adopting virtual design technology, they reduced "design-to-sample" time by 50%.

• The shift: Instead of creating all value internally, firms like Amazon and Microsoft orchestrate value created by external partners and users.

• The management change: Move from "controlling" to "enabling". Stop asking "How do we make money?" and start asking "How do we help others create value?".

• The problem: Many companies have "competing versions of the truth"—multiple systems that don't talk to each other.

• The fix: Managers must ensure a "single source of truth". AI models are only as good as the fuel you give them; bad data stymies even the best models.

• Democratize data: Train frontline workers to use data themselves to drive innovation.

• Hire "Purple People": You need talent that can bridge the gap—people with the "people skills" to work alongside technical experts.