This week we look at how AI impacts board room conversations with David Spindler, CEO of Ag Opportunities and former VP of Strategy at Tough Mudder.
As an entrepreneur, investor, and fractional CFO, David Spindler is used to operating at the cutting edge of finance management. Nevertheless, when I ask him about the biggest impact of AI on finance teams, he gives an answer grounded in common sense rather than future gazing. “Workflow optimisation.”
That’s right, workflow optimisation. Not exactly sex, drugs and rock n’ roll, or even ‘the robots are taking over’. But it’s an interesting answer when you dive into it.
“It’s an abstract answer,” admits David. “But things that may have seemed very boring at the ExCo or board level, like payments processes, are now in the spotlight because of AI. The kind of work that’s ripe for automation is front and centre. Business leaders are looking for practical applications that deliver real dollar and cents impact within the next six months.”
It’s not that any of these iterative processes are necessarily transformative by themselves, but rather that the same methods might be repeatable across different areas of a business. And also, as David warns, “if you’re not doing it, your competitor’s doing it.”
Do investors care?
As an investor, David takes a keen interest in how businesses use AI. One of the key questions he’s asking founders at present is where the clustered compute is*. “It’s almost like a raw material,” he says. For instance, if you invest in an app that relies on AI, and it’s dependent on an AI platform for its data and reasoning, what happens when prices are raised?
“In theory you can switch between platforms,” David says, “but at some point it’s going to be much more expensive. Everyone’s seen how much money is being burned and that can’t go on forever. At some point providers will be charging meaningfully more for the compute power they’re using, and what happens when the price goes up?”
David likens it to his experience at Tough Mudder, where a Facebook page had connected the brand with its community of three million people. The Tough Mudder team could rely on this platform to send free messages every day, at least until Facebook changed its monetisation policy and forced them to pay. While this didn’t stop Tough Mudder from using the platform, and an increase in LLM costs may not be a disaster for an AI-reliant company, David says it’s important for companies reliant on AI tools to understand how increasing compute costs could impact the bottom line.
Using AI the Right Way
Given that David operates in CFO functions on occasion, I ask his advice on how finance teams should use AI. He advises that teams use AI to connect different data sources and make connections between things where true APIs don’t exist. At a small scale, this might include bolting things like expense management or payroll onto Xero, QuickBooks, or a small scale ERP system, to give you the sort of customisation that previously didn’t exist.
The aim of the game here is to automate what you can, and give people more time to do the work that only people can do. Where possible and appropriate, David suggests automating processes with agentic workflows.
A Word of Caution
I ask David if anything else springs to mind when thinking about AI, and he raises one of the biggest topics of the year: fraud. This is a huge problem – the Global Anti-Scam Alliance reported that $1 trillion was lost to fraud, globally, in 2024.
Not all scam attempts are deep fakes, but AI is undoubtedly making it easier for fraudsters to operate at speed and scale with convincing attacks. “It’s much easier to build a website than it was five years ago,” says Spindler. “ It’s also much easier to build a website scam.” The same efficiencies finding their way into tech have found their way into fraud.
“Hopefully the vast majority of the impact of AI is positive, but there are some negative impacts as well that are, frankly, the same forces.”
To conclude…
There are several lessons to be taken from this. Automate what you can and consider enhancing your existing tools where possible. And if you’re a business with a proposition reliant on AI, make sure that you’re protected against, or have at least priced in, rising platform prices.
*Clustered compute refers to multiple computers working together to deliver an output. In this scenario, it refers to the potential for cloud services providers to raise their prices.
