“Imagine you are holding a glass of mud. Adding more water doesn’t make it drinkable; it just creates a bigger mess”.
The narrative that AI will replace thousands of jobs is often a convenient excuse for poor management. Ryan and I challenge the narrative of mass AI-driven layoffs. He observes that in mid-sized and smaller businesses, the focus is increasingly on repurposing talent. If you have an employee doing a repetitive, low-value task, don’t fire them—give them the tools to do something that actually drives revenue or improves customer retention.
Many organizations are falling into the same traps they faced during the digital transformation era a decade ago, and fixing the tools and processes, but not looking at what it means for humans and human systems. When we treat humans as a cost to be eliminated rather than an asset to be upskilled, we lose the institutional knowledge that keeps a business alive. True ROI comes from augmenting human capabilities, not erasing them.
Successful AI implementation depends on returning to basic project management and strategic planning. Businesses often struggle because they treat AI as a “shiny object” rather than a tool that requires specific goals. Ryan highlights that the most effective approach begins with identifying quick wins and long-term objectives before selecting any technology. Proper governance, including data auditing and clear project requirements, must precede deployment to ensure the system delivers a positive return on investment.
Does your organization have a plan to reskill team members, or are you just looking at the headcount & productivity figures?
The main insights you’ll get from this episode are :
- Clients can’t make decisions quickly enough during AI rollouts to keep up with AI and often don’t have a plan due to shiny object syndrome.
- AI adoption requires a plan first, tools later – human-driven problems hold things up; strategy and context must precede upskilling and training.
- HITL should be everywhere or wherever needed to avoid AI drift – process(es) need(s) an owner, accountability and a feedback loop.
- An AI governance team must determine the requirements for a project and work backwards to make sure the criteria are in place.
- New jobs are cropping up, e.g. AI Officer, and rollout involves some trial and error to find bespoke solutions and bring value to a company.
- Collective budgeting/accountability models do not yet represent real collaboration; structure and authority must be in place.
- Adoption in large organisations is more difficult (logistics) and depends on the industry; small companies can factor in the human element more easily.
- The quiet cost of AI is leadership identity; AI leadership is not competent at present, so good employees need to be involved rather than the leaders.
- A company needs a reason to have AI, with measurables and goals – preparation is key, i.e. seeking help (early on) and conducting research.
- The biggest blind spot at C-Suite level is relying on headlines and listening to hearsay instead of thinking things through, resulting in rushed/failed projects.
- Despite a dearth of data, immediate future will likely see job losses, but new roles – and opportunities – will be created that bring more satisfaction.
- We must go back to basics and logic, using the smartest individuals who know what is going on inside the company to involve employees at an early stage.
- Constant experimentation will increase ROI and generate positive experiences.
Find out more about Ryan and his work here :
