Steve Nolan, CEO of Ariel Group, describes a pattern he keeps seeing at large companies. Money goes into AI, training and leadership development gets cut, and layoffs get announced, all in the same breath. Steve thinks the sequence is backwards. “[After AI implementation is] when you amp [training] up to keep the people that you need,” he says.
Steve spent five years as chief revenue officer at The Myers-Briggs Company and now leads Ariel Group, a leadership development and communication training firm founded by performing artists, whose clients number a fair few of the Fortune 500. His argument rests on three things: how training budgets work today, what training returns, and what companies miss when they roll out AI without a goal.
Why Training Is the Easy Cut
When Steve first started, companies set an annual training budget and asked what could fit inside it. Over the last 10 to 15 years, those dedicated budgets have mostly disappeared. Training is now discretionary spending, funded out of individual business units, which makes it the first place to look when AI spending starts draining cash. Even a funded line is at risk. His practical tip for anyone building a budget is to use a weird number, because a round figure invites someone to announce they just saved millions.
What the Cut Costs
Steve puts the return on training far above what a cost-center view allows. “The value in my mind is 10, 20x. It’s not 1 or 2x,” he says. The case behind it is about change. Change management is the hardest thing companies are navigating, and the problems they’re navigating now often look nothing like ones they’ve had in the past. Organizations need to keep their best people to thrive during periods of immense change, and those people need training at regular intervals, not a budget freeze.
AI Without a Goal Is Just Spending
The AI rollouts carry a second problem. Employees are told to use the licenses the company has paid for, so they point them at emails and business plans with no target in mind. Steve compares it to buying 400 gallons of olives at Costco and ordering everyone to finish them before they spoil. The purchase drives the behavior, but nobody has defined the problem it was bought to solve.
His advice to learning and development leaders is this: ask leadership what the goal is and work backwards from it. Pick something concrete (Steve uses employee retention as his example) and survey the people involved. Then challenge the assumptions that come back. Skipping that step has a cost, because employees pushed to use tools without direction lose productivity. “We’re operating out of fear. We’re not operating out of opportunity,” Steve says. He admits that asking takes courage, particularly when a company is laying off thousands of people while spending ten times as much on AI. He is blunt about the stakes. If a company’s people are thinking about what their jobs will look like in a few years, he considers the company failing and its culture broken.
The Pressure Is Real. The Alternative Is Specific.
Steve does not argue that cost pressure is illegitimate. His own firm has a private equity partner; he is expected to deliver an earnings number and a revenue number in that order, and he says doing more with less is the story everywhere. But he also sees some companies reconsidering. A few are admitting they overshot on AI and are handing more authority back to their managers.
His own method for facing financial scrutiny is worth borrowing. He brings his investors data instead of hope: here is what happened, here is what it means, here are three options, and here is the one he recommends. He treats results as neither good news nor bad news, just news. A training leader defending a budget can use the same structure.
Where he sees the strongest case for training spend is cohort-based programs built around a live business problem. For example, he talks about one company that has just acquired another, with half the team coming from each side and two or three big projects ahead. Instead of sending high-potential employees through individual, asynchronous courses, he would put them together, give them the projects, pressure-test them, and then work out what support each person needs to reach the next level. Participants come away understanding their role and their colleagues better, but Steve calls that a byproduct. The aim is to help them deliver business impact.
For an executive weighing the next cut or the next license, the useful question is the one Steve keeps asking: what are we trying to achieve? A company that can answer it can defend both its AI spend and its training budget. One that cannot is stockpiling olives.
