These days, it seems every conversation eventually turns to AI. And those conversations often focus on the negative: machines taking over, jobs disappearing, and the environmental cost of all those data centers.
Meanwhile, many of us are quietly using these tools every day. Small businesses are using AI to:
- Write faster: drafting emails, proposals, marketing copy, and social media posts.
- Summarize: condensing long documents, contracts and meeting notes into the key points.
- Serve customers: answering common questions through website chat after hours.
- Handle the back-office: helping with spreadsheets, invoices, scheduling, and job descriptions.
- Brainstorm: talking through ideas, names, pricing approaches and plans.
We’ve also all heard the warnings. Be careful with the “free” versions. Don’t treat an answer as gospel, however confidently it’s delivered.
Why AI gets it wrong
I’ve tested several popular AI platforms myself, and I’ve received answers that were flat-out wrong. When I point out the error, the response is usually something like, “Good catch, sorry about that!” That’s fine when you check the answer. But not everyone does, and an unchecked mistake can be very costly.
Two issues are worth keeping in mind.
AI predicts; it doesn’t fact-check. These tools generate the answer that sounds most likely, based on patterns. Some newer tools can search the web and cite sources, but those features don’t guarantee accuracy. Sometimes it’s a confident, well-written fabrication, and the tool often sounds just as sure either way.
Some trusted sources are increasingly off-limits. Many publishers and companies now block AI tools from their content, though others have struck licensing deals that give AI companies paid access to their work. Their reasons make sense. They want to protect their intellectual property and their revenue. But when trusted sources are off-limits, AI may fill the gap from whatever it can reach, and that material isn’t always as credible.
Haven’t we seen this before?
Publishers blocking AI is only part of the picture. Advertising is starting to appear on AI platforms, and the companies behind them are under growing pressure to make money. If that sounds familiar, think back to the late 1990s and early 2000s.
The early web was a free-for-all. Netscape and Internet Explorer fought the browser wars.
WebCrawler, AltaVista, Yahoo, and AOL competed to be our front door to the internet, until Google won. Consumer content and search were largely free, and the business models were still being worked out. As advertising matured, it changed everything.
Newspapers and magazines were among the hardest hit. They put their content online for free, expecting advertising to follow. Instead, classified ads moved to free sites like Craigslist, and advertisers followed readers to Google and Facebook, which could target ads far more precisely than a printed page. Many local papers shrank, merged, or closed altogether. Many publishers learned a hard lesson about giving their work away, and it’s one reason they’re so quick to block AI today — or to demand payment for it.
Today we live with the result: pop-ups, paywalls, and ads that follow us around for weeks after a single search. AI is at that same early stage right now, with generous free access. History suggests it won’t stay there.
Follow the money
Building and running AI is extraordinarily expensive. These companies are spending billions on data centers, chips, and electricity, and the labs behind the most popular tools have yet to turn a profit. That money has to come back somehow. How it comes back will likely shape what AI becomes.
No one knows exactly how this will play out, but if the internet is any guide, here’s what we might expect.
Ads could start shaping answers. Early search engines ranked results without payment, but paid placement had arrived by the late 1990s. Over time, the top of the page filled with paid placements, and it got harder to tell an ad from an answer. AI could follow the same path. Ask for the best accounting software or a good local contractor, and someday the recommendation may go to whoever paid for it.
The gap between free and paid is likely to grow. Free users may see more ads, older models, and tighter usage limits, while paying customers get the better tools. As the old saying goes, if you’re not paying for the product, you may be the product.
A few giants may end up dominating. The early internet had dozens of search engines. Today one company handles most searches. Running AI at scale is so costly that the field could consolidate even faster, and fewer competitors usually means higher prices and fewer choices.
Quality content may come with a price tag. Publishers are already signing licensing deals with AI companies. Tools that pay for reliable sources may give better answers, and those costs could be passed on to users.
Today’s prices may be an introductory offer. Many AI tools are inexpensive today. Once businesses come to depend on them, prices, terms, and features could change. We’ve seen this with software subscriptions, and AI may be no different.
None of this means AI is a bad deal. It means the deal is likely to change. The businesses that do best will understand what they’re getting, what they’re giving up, and who is really paying for the “free” part.
What this means for your business
AI is a genuinely useful tool, but it’s no substitute for critical thinking. Treat it like a team of interns: quick, eager, and often helpful, but everything they produce needs a careful review. Double-check anything involving money, law, or customer commitments. Keep sensitive client and financial information out of tools that aren’t approved for it — check the privacy terms before you upload anything confidential, because paying for a subscription doesn’t by itself make it safe.
Pay attention to how a tool makes its money, and don’t build a critical process around something you can’t replace. Most of all, keep a person in the loop: AI can draft it, but a person should own it.
The internet didn’t destroy small business. It rewarded the ones who learned to use it wisely. AI may be no different.
Contributed by Joe Zeliger, PBC Coach
A veteran business owner himself, Joe Zeliger has over 25 years of business experience. He brings a unique understanding of what it takes to successfully found, build, and ultimately exit a highly profitable company. He now works with owners to improve profitability, strengthen teams, and increase the overall value of their businesses.
Joe welcomes the opportunity to connect and can be reached directly at zeliger@professionalbizcoach.com.
Connect with us on social media and be part of the dialog.
