The news is scary. It’s grim. There are near-apocalyptic prophecies daily about what AI will do to the world, to work, to relationships. I’m not a prophet. But I want to offer a reality check.
You might ask: does this guy actually know anything about AI’s capabilities? Fair question. I’m not a technologist. I’m not a futurist. I work closely with AI, I code with it, I build with it. But more than anything, I’m curious. I surround myself with people who know far more than I do, including business partners in AI projects I’m working on personally.
From that vantage point, I think we’re getting two things wrong at once: we’re underestimating how fast AI will commercialize inside the enterprise, and we’re overestimating the idea that the entire workforce is about to be replaced.
We’re underestimating commercialization
I believe, very soon, there will be massive commercialization opportunities for AI at the enterprise level. I’ve seen the capabilities, and they are vast. I’m a full believer that AI will replace most of the conventional software and legacy systems companies rely on today.
Legacy software companies are at a pivotal moment. I’m not telling any of them something they don’t already know. But if I ran a legacy systems company, in insurance or anywhere else, I’d be moving aggressively to find my place in the new world of software, not defending the old one.
Think about the phrase “tech stack.” Anyone in tech or operations hears it constantly. That vernacular is going to sound like Latin before long. There won’t really be a “tech stack” in the near future. This is a precarious spot for AI companies too, not just legacy ones. Many are building on top of legacy systems instead of replacing them, which only increases the total tech spend companies are pouring into AI rather than reducing it.
Efficiency isn’t the same as impact
Here’s the issue we’re dealing with in insurance right now, and I suspect most industries are facing a version of it.
AI is clearly efficient. It lets people do a lot more work in a lot less time. But what does that actually mean for the company? Is it accretive? I’d argue: not yet. Efficiency isn’t accretive until it leads to either more sales or lower expenses. It’s great that employees are getting more done. I love that. It helps on a micro level. But does it move margins, sales, or multiples? Not on its own.
So what has to happen?
The prophecy: displacement, not collapse
These tools are coming. They will replace conventional systems. They will be far more efficient. For every action there’s a reaction, especially as companies look to fund all this “efficiency.” People will be displaced. There’s no question about that.
But that’s been true for the entire span of the industrialized world. A technology arrives. People get displaced. When the combustion engine was invented, it displaced field workers, stable cleaners, horse trainers. That didn’t mean employment collapsed. It shifted. New jobs appeared: mechanics, parts manufacturers, tire makers, recyclers.
I expect the same pattern in white collar work, and insurance won’t be an exception. It’s not just that legacy insurance platforms get replaced. The AI that replaces them won’t just take over data storage, it will take over manual retrieval, processing, back-and-forth with customers, certificate issuance, routine email responses, and more. That has to happen, because without real expense reduction, there is no real efficiency gain.
What AI won’t replace, at least not in the near term, is the sales funnel itself. If anything, it will make relationships, with both carriers and clients, more important. The manual tasks inside any business will likely be heavily disrupted and largely replaced by AI systems built to displace legacy software. But replacing those jobs doesn’t mean employment falls. It’s far more likely that employment numbers rise than fall, just in a different capacity.
Why this has to work
Here’s the counterbalance. The entire economy has been built around AI for the last four years: equity markets, debt markets, data center construction, and everything else tied to the revolution driven by OpenAI and Anthropic. If massive efficiency gains don’t materialize, we have a much bigger problem than job displacement. It would mean entire industries went from huge cash flow to negative cash flow for no reason. Those are too big to fail numbers.
These are the stocks driving the economy: Microsoft, Meta, Amazon, Google, everyone with cloud compute power is betting the farm. Literally hundreds of billions of dollars spent on GPUs, on compute. We need this vision to materialize, or the fallout goes well beyond displaced jobs.
Frothy, not doomed
I know this reads more doom and gloom than optimistic. I don’t have a crystal ball. I was around for 2008, and this doesn’t feel the same. But it rhymes a little. If it’s not a bubble, I don’t know what you’d call it, but it’s certainly frothy.
And that’s really the tension at the center of everything above. The commercialization is real and coming faster than most people think, which is exactly why the money pouring into this makes sense. But that same money is what turns “faster than expected” into “has to happen, or else.” Hundreds of billions in compute doesn’t get spent on a maybe. The bet only works if efficiency actually shows up on the income statement, and efficiency showing up on the income statement is what displaces jobs. The upside case and the disruption case are the same case. You don’t get one without the other.
So no, I don’t think this ends in mass unemployment. I think it ends in a lot of shifting, the way it always has. What I’m less sure about is the timeline, and how bumpy the transition gets before the new jobs show up to replace the old ones. This is the piece that will be stressful. Our greatest fears lie in anticipation. But Americans have always been resilient. We aren’t the envy of capital markets everywhere else around the world because we’re the smartest. It’s because we take risk. And we put our money where our mouth is... and to work. We also adapt quickly. Silicon Valley-like adventurism doesn’t exist anywhere else in the world because nobody bets bold like we do. That’s not a political statement. It’s a statement about our people and their drive.
I choose to see this as an opportunity. I’m curious by nature, and that curiosity makes me look for opportunity around every corner. I don’t think this changes the economic disparity between the haves and have nots, if anything it’s likely to widen it. We need government help for that one. But for those willing to adapt, embrace, and actually understand AI, this could be one of the more opportunistic and positive stretches ahead.
Thanks to our sponsors: Freeflow - Freeflow.ai and Imperial PFS Imperial PFS !
RiskCellar is part of the Max Revenue Media Network.
