Business AI is not an Individual Responsibility
When I meet people, I make a point of asking them if and how they're using AI at work. The answers are diverse and span the whole spectrum
Some say that they're using ChatGPT all the time to refine their writing, create firsts drafts, or conduct initial research
Others lament that the only tool they're allowed to use is Microsoft's Copilot, in a "sandboxed" way where Copilot doesn't even get access to company documents, and that they find it underwhelming
What all the answers I've gotten so far have in common is that none point toward strategic, organization-wide rollouts of AI. At best, AI is seen as something the individual worker must incorporate into their workflow, all while the workers themselves are part of a company-wide workflow that lacks any intentionality regarding AI.
This is nuts!
Would Henry Ford have been able to massively ramp up the production of Model T's by focusing only on the productivity and efficiency of each individual worker? No. He needed to reimagine the entire way cars were being built. The assembly line wasn't something that each worker adopted (or not). It was something introduced at the company level and required organizing every aspect of the business around it.
Why should it be any different with AI? If we just hand our employees the reins to ChatGPT (or Microsoft Copilot), the gains will be marginal at best. Adoption will be haphazard. Worse, you might just incentivize everyone to create copious amounts of workslop.
It doesn't have to be like that, and it all comes down to a few common-sense principles:
First, understand your current processes. How does work get done? Where are the points of friction? Where are the bottlenecks? What is being done that shouldn't be done, what isn't being done that should be done?
Then, understand where and how AI could help. Some examples:
AI can reduce the friction of manual data wrangling. This is great for synthesizing reports (e.g. board decks) from disparate sources without having to interrupt workers with these requests for information
AI can efficiently categorize, filter, and route information. Valuable team members spending too much time sifting through unstructured data, tagging requests or deciding which department an incoming request should go to? That's a job for AI.
AI can find the needle in a haystack of messy incoming data: Receiving opportunities (say, pitch decks from startups when you're an investor) in messy formats and need to find, extract, and match on your criteria? AI will do a good job.
Finally, identify a new AI-empowered workflow: How does it integrate with the rest of the business? What has to change in the way work gets done based on this workflow? What does it mean for the team members involved in that work?
Once you have clarity on that, you can build the AI solution and roll it out, confident that it'll make a measurable impact on your operations (you do have KPIs or measurement targets in place, right?).
