I think we’ve got this whole AI thing backwards.
Every Monday morning, my inbox is a graveyard of "must-see" updates. This week is no different. We’ve seen OpenAI nudging their models towards better reasoning, Google tinkering with Gemini’s integration into everything but the kitchen sink, and a handful of niche tools promising to automate away our remaining brain cells.
The natural instinct is to rush in. We feel this nagging guilt if we aren't "leveraging" (ugh, sorry) the latest release within twenty minutes of it hitting X.
But after spending the week looking at what’s actually landed, I’m convinced that for most businesses, the smarter move is to ignore about 90% of it. Most of these updates are like buying a high-performance Ferrari to drive to the local corner shop. It’s impressive, sure, but the pedestrian crossing doesn't care about your 0-60 time.
Why OpenAI’s Latest Updates Might Be Overkill
The talk of the town this week has been the continued refinement of "reasoning" models. The idea is that the AI takes a beat, thinks about the problem, and then gives you an answer that isn't complete nonsense.
In a laboratory setting, this is magnificent. For a developer trying to stitch together a particularly stubborn bit of Python, it’s a godsend. But for the average business owner trying to improve their sales follow-up or automate a bit of data entry? It’s often unnecessary.
Most business problems aren't complex logic puzzles; they’re execution puzzles.
If you are using AI to draft an email based on a CRM entry, you don't need a model that can pass the Bar Exam. You need a model that knows how to sound like a human who’s had a decent cup of coffee. The "reasoning" models are slower and often more expensive. Unless you’re solving genuinely hard mathematical or structural problems, your current setup is probably fine.
Actually, it’s better than fine. It’s stable. And stability is the one thing no one mentions in the AI hype cycle.
Google Gemini and the Problem of "Everywhere-ness"
Google has been busy pushing Gemini deeper into the Workspace suite. The promise is that the AI will be there, hovering like a helpful ghost, while you write your docs and check your emails.
It sounds efficient. In practice, it’s often just another distraction.
I’ve noticed a pattern with these "integrated" AI features: they solve the wrong problem. Most people don't struggle to write the words in a document. They struggle to decide what the document should say in the first place.
Having an AI offer to "summarise this thread" or "generate a table" feels like progress, but it’s often just a way to avoid doing the actual thinking. It’s the digital equivalent of tidying your desk instead of doing your taxes. You feel busy, but the needle hasn't moved.
If you’re running a team, the challenge isn't giving them more tools to generate text. The challenge is ensuring they know how to judge if that text is actually rubbish.
Digital Fatigue and the Search for Value
I had a chat with a client yesterday who was stressed because they hadn't yet implemented a custom GPT for their customer service. They felt they were "falling behind" the competition.
I asked them how their current customer service was doing. "Oh, it's great," they said. "We reply within ten minutes and our customers love us."
Why on earth would you risk that reputation by sticking an unproven bot in the middle of it just to keep up with the Joneses?
We are seeing a strange phenomenon where businesses are willing to sacrifice human connection—the very thing that makes them valuable—for a slight increase in "efficiency" that no one actually asked for.
If you want to stay ahead, stop looking at the feature list and start looking at the friction in your business.
- Is your sales process clunky because you’re slow to respond? Then automate the alert, not the conversation.
- Are you losing leads because you don't follow up? Then use a simple sequence, not a complex AI agent.
- Are you spending eight hours a week on spreadsheets? fine, let the AI have a go at that.
You can find more articles on AI that focus on the practical side of these tools rather than the shiny wrappers.
How to Filter AI News Without Losing Your Mind
If you’re feeling overwhelmed by the constant stream of "game-changing" (there's that word again) releases, try this filter:
- Does this solve a problem I actually have? If you weren't complaining about it last Tuesday, you probably don't need a solution for it this Friday.
- Does it make the customer experience better or just cheaper for me? If it’s only the latter, be very careful. Customers can smell "cheap" from a mile away.
- Is the "old" way actually broken? Sometimes a simple checklist is more effective than a neural network. It's certainly easier to fix when it goes wrong.
- Can I explain what it does to my nan? If you can’t describe the benefit in plain English, it’s likely just hype.
Most of the time, the best way to use AI is to wait three months. Let the early adopters find the bugs, let the pricing stabilise, and let the "revolutionary" features fail. The ones that are still standing after ninety days are the ones worth your time.
The Practical Path Forward
What should you actually do this week?
Instead of trying five new tools, pick one process in your business that feels heavy. Maybe it's how you handle initial enquiries, or how you prep for meetings.
Take a look at your existing tools. Most of the time, the software you already pay for has added an AI feature that you’ve ignored. Start there. It’s already in your workflow, it’s already paid for, and it doesn't require a new login.
If you’re trying to figure out how to navigate this without falling into the hype trap, feel free to book a consultation. We can talk about what actually works and what is just expensive window dressing.
And if it's Thursday, you really ought to get the free book which goes into detail on how to build a business that relies on smart systems, not just the latest shiny object.
AI is a tool, not a strategy. It’s a hammer, not the house. This week, let everyone else marvel at the new, gold-plated hammers hitting the market. You just focus on the building.
The irony of the AI revolution is that as the technology becomes more common, the value of human judgment, wit, and genuine connection goes through the roof. If you can use AI to clear the boring stuff off your plate so you can be more human, you’ve won. If you use it to replace the human parts, you’re just a very expensive calculator.
I cover the latest AI releases and tools every Friday at steventann.com. If you found this useful, there's plenty more where this came from.
About the Author
Steven Tann is an AI consultant, author of "You're Selling AI Wrong", and founder of SalesM8. He writes about AI, sales, and running a business from a narrowboat on the English canals. Connect with him at steventann.com.
