By Dustin Leiblein
I think a lot of businesses are underestimating AI because their experience with it hasn't caught up with what the tools can actually do.
They've tried a chatbot. They've had it write an email or a social post. Maybe the result was useful. Maybe it sounded generic. Either way, that becomes their understanding of AI.
Meanwhile, people who spend time with more capable models are learning how to use them for increasingly difficult work. They're researching, building, testing, and solving problems they previously needed more time or help to work through.
That's a real advantage. And I think the gap between those two experiences has room to grow.
It reminds me of the early internet
The early internet gave businesses new ways to reach customers and get work done. Businesses that figured out how to use it had opportunities that others hadn't even started exploring.
AI reminds me of that period. Except this time, it's harder to see how far apart two businesses actually are.
You can look at a website and get some idea of what a company built. With AI, two people can sit in front of nearly identical text boxes and do completely different levels of work.
One asks for a caption. The other uses it to investigate why a campaign is attracting the wrong leads, question the explanation, and work out what evidence would settle it.
Both say they're using AI. That doesn't mean they're getting anything close to the same value.
Access is the first step
I understand why someone would question paying for AI if all they've used it for is rewriting a paragraph. The value of better access becomes clearer when you start asking more of the tool.
A more capable model can matter. So can having enough capacity to use it throughout your work. But paying for access doesn't automatically give you the skills to use it well.
You still have to understand the problem. You have to give it useful context. And you have to know enough to recognize when the answer is wrong, incomplete, or just plausible.
That last part matters a lot. AI can give you a confident answer to a question you haven't thought through. If you can't evaluate the answer, you can walk away feeling productive without having solved anything.
The person using it still matters
I think the human side of this gets overlooked because the technology does so much of the visible work.
Someone still decides what to investigate, what evidence matters, and what a good result looks like. Those decisions shape everything that follows.
If you understand marketing, you can push past a generic campaign recommendation. If you understand how a website works, you can challenge a fix that addresses the symptom while leaving the cause in place.
AI gives you more room to work with that knowledge.
You can also use it to build knowledge you don't have yet. Research on customer support found that novice and lower-performing workers gained the most from an AI assistant. That is a useful reminder that experience isn't a prerequisite for benefiting from the technology. Generative AI at Work, NBER.
The willingness to learn and check your own thinking matters. Having a powerful tool available doesn't guarantee you'll do either.
The advantage can keep building
This is the part I think businesses should pay attention to.
When you use AI seriously, you can learn more than how to finish one task faster. You learn how to explain problems clearly, break difficult work into manageable pieces, test an answer, and recognize which parts still need your judgment.
Those skills carry into the next task. The useful context and processes you develop can carry forward too.
Then a better model arrives, and you already have work you can test it against. You can find out where the improvement actually helps.
A business that keeps putting this off still has that learning ahead of it. Getting access later doesn't erase the practice someone else has already put in.
I don't expect access alone to close that gap. The tools can become easier to use while the difference in how people use them remains substantial.
The lead isn't guaranteed or permanent. But waiting has a cost, and part of that cost is the learning you're delaying.
You still have to check the work
Being enthusiastic about AI doesn't mean I think it should be trusted without question.
Research with Boston Consulting Group consultants found that AI improved performance on tasks within its capabilities. On a task outside those capabilities, people using AI were less likely to reach the correct answer. Harvard Business School working paper.
That's why I think verification has to be part of using it well. Check the source. Test the change. Compare the recommendation with what is actually happening in the business.
Moving faster is useful when you're making progress. It's also a very efficient way to spread a mistake.
Start using it on a problem that matters
My advice is to pick one recurring problem in your business and start there.
Give AI the context. Work through the problem with it. Challenge the answer. Check whether the result improves the work. Then take what you learned into the next attempt.
That's the approach I want AI to support at Squid X Media: better work, with evidence behind it and a person responsible for the outcome.
I think the businesses that keep doing this are putting themselves in a strong position. They're building experience while others are still judging the technology by the first thing they tried.
If you're trying to work out where AI could help your marketing or website, book a free 30-minute strategy consultation. Bring the problem. We can help you find a useful place to start.
Leave a Reply