How Small Businesses Can Unlock AI’s Potential
· Peter Hall
Running a business takes a surprisingly wide range of skills, from accounting to product design to market research. Big companies can hire specialists for each role.
Running a business takes a surprisingly wide range of skills, from accounting to product design to market research. Big companies can hire specialists for each role. But for small and medium sized businesses, that’s often not possible. That’s where large language models like GPT-4 come in, not as a replacement for humans, but as a flexible tool that can handle many different tasks without needing a dozen new hires.
The MIT Technology Review piece explores how these AI systems are being tested across industries that can’t afford deep expertise in every area. Instead of training a separate model for tax prep and another for logo design, a single LLM can be prompted to help with both. The trick is learning how to talk to it clearly and break down problems into steps the model can follow. Think of it as hiring a very fast, very polite intern who knows a little bit about everything.
One real world example the article highlights: a small accounting firm used an LLM to draft client emails and summarize tax code changes. The firm’s owner said it saved hours each week without sacrificing accuracy. The catch is that you still need someone who knows what they’re doing to check the output. The AI doesn’t replace judgment, it just speeds up the grunt work.
So what’s next? The real opportunity might not be in building smarter models, but in building smarter workflows around the ones we already have. The companies that figure out how to weave these tools into their daily operations, without chasing hype, will likely end up with a real edge. And the rest? They’ll be stuck trying to hire that unicorn expert who can do everything.