Latest / Ctrl AI Profit / Ep. 111 | The Federal Reserve Just Weighed In on AI
Transcript
- 0:20SPEAKER_00: The Federal Reserve just published a research piece on AI's macroeconomic challenges and promises.
- 0:26SPEAKER_00: And for once, it's not hype.
- 0:29SPEAKER_00: It's the most honest assessment of AI's economic impact I've seen from a major institution.
- 0:35SPEAKER_01: The New York Fed's Liberty Street Economics blog published an analysis that looks at AI through the lens of productivity, labor, inflation, and financial stability.
- 0:46SPEAKER_01: And what they found is nuanced in a way that most AI coverage isn't.
- 0:49SPEAKER_01: AI isn't going to save the economy overnight, and it's not going to destroy it either.
- 0:54SPEAKER_01: The truth is somewhere in the middle, and it depends almost entirely on how businesses respond.
- 1:01SPEAKER_00: Let's start with productivity because that's the big promise everyone makes about AI.
- 1:06SPEAKER_00: The feds take AI could boost productivity significantly, but the effect is uneven and slow to materialize.
- 1:14SPEAKER_00: Sound familiar?
- 1:15SPEAKER_01: It should.
- 1:16SPEAKER_01: We've been saying this on the show for months.
- 1:19SPEAKER_01: AI doesn't automatically make you more productive, it gives you the capacity to be more productive.
- 1:24SPEAKER_01: Whether you actually are depends on what you do with it.
- 1:28SPEAKER_01: The Fed is now confirming this at a macroeconomic level.
- 1:32SPEAKER_00: The Fed researchers point out that productivity gains from AI are concentrated in specific sectors: information technology, professional services, finance, while other sectors like construction, hospitality, and healthcare see much smaller gains.
- 1:47SPEAKER_00: That's not a uniform productivity revolution.
- 1:50SPEAKER_00: That's a sector-by-sector transformation.
- 1:54SPEAKER_01: And that has real implications for small businesses.
- 1:57SPEAKER_01: If you're in a sector that's seeing big AI productivity gains, you need to adopt fast or you'll be outpaced.
- 2:03SPEAKER_01: If you're in a sector that's seeing modest gains, you have more time, but also less room for error because your margins are thinner.
- 2:11SPEAKER_00: The second big finding is about labor.
- 2:14SPEAKER_00: The Fed says AI will displace some jobs, create others, and reshape most.
- 2:19SPEAKER_00: But, and this is the key, the transition period is going to be messy.
- 2:24SPEAKER_00: Workers who lose jobs to AI may not have the skills for the new jobs AI creates.
- 2:29SPEAKER_00: And the timeline for new job creation is uncertain.
- 2:34SPEAKER_01: This is the part that doesn't get enough attention.
- 2:36SPEAKER_01: Everyone says AI will create new jobs, and it will.
- 2:40SPEAKER_01: But the gap between job destruction and job creation could be years.
- 2:44SPEAKER_01: And for the people in that gap, it's not theoretical.
- 2:48SPEAKER_01: It's their livelihood.
- 2:50SPEAKER_00: For small business owners, the labor implication is twofold.
- 2:54SPEAKER_00: First, if you're in an industry where AI is displacing workers, you might find a pool of available talent at lower wages.
- 3:02SPEAKER_00: That's an opportunity.
- 3:03SPEAKER_00: Second, if you're trying to hire in an industry where AI is creating new roles, you'll face wage pressure because the skills are scarce.
- 3:12SPEAKER_00: That's a challenge.
- 3:14SPEAKER_01: The Fed also highlights something that doesn't get talked about enough.
- 3:17SPEAKER_01: AI could increase inequality.
- 3:20SPEAKER_01: Workers with AI skills will command premium wages.
- 3:24SPEAKER_01: Workers without them will face downward pressure.
- 3:26SPEAKER_01: And small businesses that can't afford to upskill their workforce will fall further behind.
- 3:31SPEAKER_00: Now let's talk about the inflation angle because this is where the Fed really gets interesting.
- 3:36SPEAKER_00: AI could reduce inflation by making businesses more efficient, lower costs, faster processes, fewer errors.
- 3:45SPEAKER_00: But it could also increase inflation in the short term by driving up demand for compute, energy, and specialized labor.
- 3:52SPEAKER_01: Think about it.
- 3:54SPEAKER_01: Every company is buying GPUs, every data center is expanding, every cloud provider is raising prices.
- 4:01SPEAKER_01: The demand for AI infrastructure is creating inflationary pressure, even as AI promises to reduce costs elsewhere.
- 4:08SPEAKER_01: It's a push-pull dynamic.
- 4:10SPEAKER_00: And for small businesses, this means two things.
- 4:13SPEAKER_00: First, the AI tools you're buying might get more expensive before they get cheaper.
- 4:17SPEAKER_00: Second, the cost savings from AI might take longer to reach you because the big companies are capturing them first.
- 4:24SPEAKER_01: The Fed's analysis acknowledges this.
- 4:26SPEAKER_01: They note that large enterprises are better positioned to capture AI productivity gains because they can invest in infrastructure and talent.
- 4:34SPEAKER_01: Small businesses have to rely on off-the-shelf tools and hope the savings trickle down.
- 4:39SPEAKER_00: But here's what the large companies can't do that small businesses can.
- 4:44SPEAKER_00: They can't move fast, they can't pivot quickly, they can't try something on Monday and change direction on Wednesday.
- 4:50SPEAKER_00: Small businesses have an agility advantage that partially offsets the scale advantage of large companies.
- 4:58SPEAKER_01: That's exactly right.
- 4:59SPEAKER_01: A small business can adopt a new AI tool in a week.
- 5:03SPEAKER_01: A large company takes six months to evaluate, approve, and deploy the same tool.
- 5:08SPEAKER_01: The small business captures the productivity gain faster.
- 5:11SPEAKER_01: The question is whether they reinvest it.
- 5:14SPEAKER_00: Which is why the reinvestment gap we talked about in our last episode is so critical.
- 5:19SPEAKER_00: If you're a small business and AI saves you time, but you don't reinvest that time into revenue generating activities, you're not just falling behind.
- 5:30SPEAKER_00: You're falling behind while also paying for the tool that's not helping you grow.
- 5:35SPEAKER_01: It's the worst of both worlds.
- 5:37SPEAKER_01: You're spending money on AI, but not capturing the value.
- 5:40SPEAKER_01: The Fed calls this a productivity paradox, more investment in technology without proportional productivity gains.
- 5:47SPEAKER_01: And it's exactly what's happening in the 72% of companies that Gartner identified.
- 5:52SPEAKER_00: The Fed's third big area is financial stability.
- 5:57SPEAKER_00: They're concerned that AI could create new systemic risks, flash crashes driven by AI trading algorithms, concentration risk from a few large AI providers, and model risk from widespread dependence on the same AI systems.
- 6:12SPEAKER_01: For small businesses, the systemic risk that matters most is concentration.
- 6:17SPEAKER_01: If your business depends on one AI provider, whether it's OpenAI, Google, or Anthropic, and that provider has an outage, a price increase, or a policy change, your business is exposed.
- 6:31SPEAKER_01: The Fed is worried about this at a macro level.
- 6:34SPEAKER_01: You should be worried about it at a micro level.
- 6:36SPEAKER_00: Let me give you a real example.
- 6:38SPEAKER_00: A few months ago, OpenAI had a major outage that lasted hours.
- 6:43SPEAKER_00: Every business that relied on Chat GPT for customer service, content generation, or data analysis was down.
- 6:50SPEAKER_00: No backup, no alternative.
- 6:52SPEAKER_00: Just stopped.
- 6:54SPEAKER_01: And that's a company with good uptime.
- 6:56SPEAKER_01: The less reliable providers have outages more frequently.
- 7:00SPEAKER_01: If your entire AI stack depends on one provider, you're one outage away from being completely offline.
- 7:06SPEAKER_01: That's not a theoretical risk.
- 7:08SPEAKER_01: That's a Tuesday afternoon problem.
- 7:10SPEAKER_00: This connects directly to what we said about diversification.
- 7:14SPEAKER_00: If all your AI tools run on OpenAI's infrastructure, you're concentrated.
- 7:21SPEAKER_00: If one tool uses OpenAI, another uses Google, and a third runs locally, you're diversified.
- 7:28SPEAKER_00: The same risk management principle that applies to your investment portfolio applies to your AI stack.
- 7:35SPEAKER_01: And the Fed is essentially saying the same thing at a macro level.
- 7:38SPEAKER_01: Concentration risk is real.
- 7:40SPEAKER_01: Too many businesses depending on too few AI providers creates systemic vulnerability.
- 7:46SPEAKER_01: The solution is the same whether you're a central bank or a small business.
- 7:50SPEAKER_01: Diversify.
- 7:51SPEAKER_00: Let me pull this all together for the small business owner.
- 7:54SPEAKER_00: The Fed says three things.
- 7:56SPEAKER_00: AI will boost productivity, but unevenly.
- 7:59SPEAKER_00: AI will reshape labor but with a messy transition.
- 8:02SPEAKER_00: And AI creates financial stability risks, especially from concentration.
- 8:07SPEAKER_01: Three Fed findings, three practical takeaways.
- 8:10SPEAKER_01: First, don't assume AI will automatically make you more productive.
- 8:15SPEAKER_01: Have a plan for reinvesting save time into revenue generating activities.
- 8:19SPEAKER_01: Second, start upskilling yourself and your team now.
- 8:22SPEAKER_01: The labor transition is coming and the workers with AI skills will have the advantage.
- 8:27SPEAKER_01: Third, diversify your AI stack.
- 8:30SPEAKER_01: Don't put all your eggs in one provider's basket.
- 8:33SPEAKER_00: I want to highlight something the Fed said that I think is underappreciated.
- 8:37SPEAKER_00: They noted that AI's macroeconomic impact depends almost entirely on adoption speed and quality, not on the technology itself, on how quickly and how well businesses adopt it.
- 8:49SPEAKER_01: Which means the variable isn't the AI.
- 8:53SPEAKER_01: It's you.
- 8:54SPEAKER_01: It's your business.
- 8:55SPEAKER_01: It's your willingness to adopt, adapt, and reinvest.
- 8:59SPEAKER_01: The technology is ready.
- 9:00SPEAKER_00: The question is whether businesses are ready to use it well.
- 9:03SPEAKER_00: And that's both encouraging and sobering.
- 9:06SPEAKER_00: Encouraging because it means the outcome is in your hands.
- 9:10SPEAKER_00: You're not passive.
- 9:12SPEAKER_00: You can choose to adopt well, reinvest wisely, and grow.
- 9:16SPEAKER_00: Sobering because it means there's no shortcut.
- 9:19SPEAKER_00: You have to do the work.
- 9:21SPEAKER_01: The Fed's not going to tell you how to run your business, but they're telling you that the businesses that adopt AI well will pull ahead, and the ones that don't will fall behind.
- 9:31SPEAKER_01: That's not hype.
- 9:32SPEAKER_01: That's economics.
- 9:34SPEAKER_00: Let me make this really concrete.
- 9:36SPEAKER_00: The Fed says AI productivity gains are concentrated in specific sectors.
- 9:42SPEAKER_00: If you're in professional services, technology, or finance, AI is already making a measurable difference.
- 9:47SPEAKER_00: If you're in construction, hospitality, or retail, the gains are smaller but coming.
- 9:52SPEAKER_00: The key in both cases is the same.
- 9:55SPEAKER_00: Adopt with a plan, not just to purchase.
- 9:58SPEAKER_01: And the Fed's data supports this.
- 10:00SPEAKER_01: The sectors seeing the biggest productivity jumps are the ones where businesses are not just buying AI tools, but redesigning their workflows around them.
- 10:08SPEAKER_01: Buying a tool is step one.
- 10:10SPEAKER_01: Redesigning your work to leverage it is steps two through ten.
- 10:14SPEAKER_00: The Fed also warns about something called model risk.
- 10:17SPEAKER_00: The danger that many businesses are using the same AI model, trained on the same data, making the same decisions.
- 10:24SPEAKER_00: If the model is wrong, everyone is wrong at the same time.
- 10:28SPEAKER_01: This is the herding problem.
- 10:30SPEAKER_01: When everyone uses the same AI to make the same kinds of decisions: pricing, hiring, lending, inventory, you get correlated errors.
- 10:41SPEAKER_01: And correlated errors can cascade into market disruptions.
- 10:45SPEAKER_00: For a small business, the herding problem looks like this.
- 10:49SPEAKER_00: If you and all your competitors are using the same AI tool to set prices, optimize routes, or target customers, you're all making similar decisions, which means the competitive advantage goes to the business that uses AI differently or supplements it with human judgment.
- 11:04SPEAKER_01: Think about it.
- 11:05SPEAKER_01: If every pest control company in town is using the same AI to optimize their routes, they're all going to end up with similar efficiency.
- 11:13SPEAKER_01: The one that differentiates is the one that adds something the AI can't, a personal call to check on a longtime customer, a handwritten thank you note, a willingness to handle the weird edge cases that AI doesn't cover.
- 11:25SPEAKER_00: That's where the human touch becomes the moat.
- 11:27SPEAKER_00: Not in replacing AI, but in complementing it.
- 11:30SPEAKER_00: AI handles the routine, you handle the relationship.
- 11:33SPEAKER_00: AI handles the efficiency, you handle the experience.
- 11:37SPEAKER_01: And customers can tell the difference.
- 11:40SPEAKER_01: They might not be able to articulate it, but they know when they're being processed by an algorithm versus being served by a person who cares about their business.
- 11:48SPEAKER_00: The Federal Reserve isn't going to tell you to write thank you notes, but they did tell you that AI's economic impact depends on how you use it.
- 11:56SPEAKER_00: And we're telling you that the best way to use it is to combine AI efficiency with human judgment.
- 12:03SPEAKER_01: Human judgment is the differentiator.
- 12:05SPEAKER_01: AI can optimize within the parameters it's given.
- 12:08SPEAKER_01: But setting the parameters, deciding what to optimize for, understanding context, reading between the lines, that's still a human job.
- 12:17SPEAKER_01: The businesses that combine AI efficiency with human judgment will outperform the ones that rely on AI alone.
- 12:25SPEAKER_00: And that's actually encouraging for small business owners, because human judgment is what you do best.
- 12:30SPEAKER_00: You know your customers, you know your market, you know the things the AI doesn't know because it can't know.
- 12:37SPEAKER_00: The context, the relationships, the gut feelings that come from decades of experience.
- 12:43SPEAKER_01: The AI can tell you that Route A is 5% faster than Route B, but you know that Mrs.
- 12:48SPEAKER_01: Johnson on Route B always needs an extra 10 minutes because she wants to talk about her garden.
- 12:53SPEAKER_01: That's the kind of judgment that creates customer loyalty.
- 12:56SPEAKER_01: And that's not in the algorithm.
- 12:58SPEAKER_00: The best businesses won't be the ones that replace humans with AI.
- 13:02SPEAKER_00: They'll be the ones that use AI to free up humans for the work that only humans can do: the relationship work, the judgment calls, the moments that make customers say, these people get me.
- 13:13SPEAKER_01: And that's actually the Fed's implicit message, too.
- 13:16SPEAKER_01: They're saying AI's impact depends on adoption quality.
- 13:20SPEAKER_00: Quality means using AI well, not just using it.
- 13:23SPEAKER_00: The Federal Reserve just published the most level-headed analysis of AI's economic impact we've seen.
- 13:30SPEAKER_00: Their conclusion: AI's promise is real, but it's not automatic.
- 13:34SPEAKER_00: The gains go to the businesses that adopt with intention, reinvest with discipline, and diversify with purpose.
- 13:41SPEAKER_01: The Fed said it, we've been saying it.
- 13:43SPEAKER_01: The businesses that treat AI as a magic bullet will be disappointed.
- 13:47SPEAKER_01: The businesses that treat it as a tool that requires strategy, reinvestment, and human judgment will thrive.
- 13:53SPEAKER_01: It's not complicated.
- 13:54SPEAKER_01: But it's not easy either.
- 13:56SPEAKER_00: It's not easy, but it is simple.
- 13:58SPEAKER_00: And simple beats complicated every time in small business.
- 14:02SPEAKER_00: You don't need a 20-page AI strategy document.
- 14:05SPEAKER_00: You need three things: a tool that saves you time, a plan for what to do at that time, and the discipline to follow through.
- 14:13SPEAKER_01: Three things.
- 14:14SPEAKER_01: That's it.
- 14:15SPEAKER_01: Tool, plan, discipline.
- 14:16SPEAKER_01: The Fed just spent how many pages confirming what every small business owner already knows in their gut.
- 14:22SPEAKER_01: The magic isn't in the technology, it's in what you do with it.
- 14:26SPEAKER_00: The productivity promise is real, the labor transition is coming, the concentration risk is real.
- 14:31SPEAKER_00: And the difference between the businesses that thrive and the ones that don't comes down to one thing: what you do with the time and capacity AI gives you.
- 14:40SPEAKER_01: Adopt with a plan.
- 14:42SPEAKER_01: Reinvest with discipline.
- 14:44SPEAKER_01: Diversify your stack.
- 14:46SPEAKER_01: The Fed just gave you the playbook.
- 14:48SPEAKER_00: Now use it.
- 14:49SPEAKER_00: The Federal Reserve just told you AI's economic impact depends on you, not on the technology.
- 14:55SPEAKER_00: On you.
- 14:56SPEAKER_00: Don't waste that insight.
- 14:58SPEAKER_00: The businesses that adopt with a plan, reinvest with discipline, and diversify with purpose will thrive.
- 15:04SPEAKER_00: The rest will wonder what happened.
- 15:06SPEAKER_00: That's the Fed's message.
- 15:07SPEAKER_00: That's our message.
- 15:09SPEAKER_00: Now it's your move.
- 15:10SPEAKER_00: See you next time on control AI Profit.