There’s a story that keeps showing up in headlines, surveys, and even graduation ceremonies: people are tired of being sold AI as the answer to every problem. And paradoxically, that’s good news for businesses that want to use it well.
Two recent data points paint the picture. On one hand, surveys from Pew and Gallup show that most Americans don’t trust AI or the people in charge of it. On the other, just a few days ago, students at the University of Arizona booed Eric Schmidt — former Google CEO — during his commencement speech when he started praising AI.
These aren’t isolated anecdotes. They’re symptoms. And if you run a business and are thinking about automating processes with AI, understanding what’s going on here can save you a lot of money and a lot of internal friction.
What’s actually happening
Let’s stick to the facts. At the graduation event, Schmidt praised AI’s capabilities in front of an audience that wasn’t in the mood for triumphalist speeches. The reaction was immediate: boos, discomfort, and plenty of headlines the next day. This wasn’t a radical group or an organized protest — it was university students reacting to a message they felt was disconnected from their reality.
And surveys back that up. When ordinary people are asked whether they trust AI or the people building it, the majority answer no. It’s not just fear of unemployment or sci-fi scenarios — it’s something more concrete: the sense that decisions about AI are being made far above them, without anyone being consulted, while the costs are paid lower down.
The problem isn’t AI. The problem is how it’s being sold and how it’s being rolled out.
Why this matters to you as a business owner
You might be thinking: “Okay, but I’m not Eric Schmidt. I just want my team to stop wasting three hours a day answering repetitive emails.” And you’re right. But public distrust of AI has direct consequences for how you implement it in your company:
- Your team will have doubts. If your employees read the same headlines as everyone else, they’ll walk into the “we’re implementing AI” meeting with their arms folded. If you don’t manage that conversation, the project fails before it starts.
- Your customers have doubts too. A badly placed chatbot, an automated email that sounds robotic, a synthetic voice call with no context — each mistake reinforces the broader distrust and reflects on your brand.
- Regulators are watching. The EU already has the AI Act. Rolling out AI without thinking about transparency and traceability is building on sand.
- The difference between “AI that works” and “AI that’s embarrassing” is becoming obvious. Customers notice the difference. And they forgive it less and less.
The hype trap: three ways businesses mess up AI
After years of implementing AI solutions in real companies, there are three mistakes we keep seeing over and over. They’re the same mistakes driving public distrust.
1. Implementing AI to look impressive, not to solve something
The question is not “how do we put AI into the business?” The right question is “what specific problem do we have, and is AI the best tool to solve it?” Sometimes it is. Sometimes it isn’t. A Python script or a solid manual process can be better than an expensive language model.
2. Replacing people instead of freeing them
When the story becomes “we’re going to cut jobs thanks to AI,” the team pushes back and customers run. When the story becomes “we’re going to take away the tasks you hate so you can do the work that adds value,” everything changes. AI should handle repetitive work. People should handle creativity, strategy, and decisions.
3. Black boxes with no transparency
An AI that makes decisions and nobody knows how is a problem waiting to explode. Not just for compliance — also because when something goes wrong (and it will go wrong at some point), you need to be able to look inside, understand what happened, and fix it.
How to implement AI well: a practical guide
All right, we’ve seen what not to do. Now let’s look at what to do. This is what we apply at Studio SmartWork in every project, and what we’d recommend to any business getting started.
Step 1: Start with the problem, not the technology
Before talking about models, agents, or any tool, answer these questions:
- What specific task is eating up your team’s time every week?
- Exactly how many hours?
- What would happen if that task disappeared tomorrow?
- Is it repetitive work or does it require human judgment?
If you don’t have clear answers, you’re not ready to implement AI. You’re ready to do a process audit first.
Step 2: Choose processes where risk is low and impact is high
Don’t start by automating the most critical decision in your business. Start with something where:
| Characteristic | Why it matters |
|---|---|
| It’s repetitive | AI is good at patterns |
| It has clear rules | It reduces ambiguity |
| Errors are reversible | You can iterate without disasters |
| It takes a lot of time | The ROI is obvious |
| Data is available | Without data, there’s no AI |
Good candidates: lead qualification, inbox management, first-line customer support, draft proposal generation, meeting scheduling.
Poor candidates to start with: hiring decisions, medical diagnoses, legal or financial decisions without supervision.
Step 3: Design with humans in the loop
AI doesn’t have to replace the human. It can prepare the work so the human can decide. Examples:
- An email comes in → AI classifies it, drafts a reply, and leaves it in the inbox → a person reviews and sends it.
- A lead comes in → AI enriches it, scores it, and assigns it → sales only calls the good ones.
- A call comes in at 10 p.m. → the voice agent collects details and books the appointment → the next day, the team confirms.
This reduces mistakes, builds trust within the team, and makes it possible to improve the system gradually.
Step 4: Measure, monitor, adjust
AI that gets deployed and forgotten is AI that breaks. You need to monitor:
- How many tasks does it process per day?
- How often does it make mistakes?
- What kinds of errors does it make?
- Is it improving or getting worse over time?
Without metrics, you don’t know if it works. And if you don’t know if it works, you can’t defend it when someone asks.
Step 5: Be transparent with your team and your customers
If a customer is talking to a bot, say so. If a decision is made by an automated system, explain it. Transparency doesn’t reduce value — it adds it. People forgive honest mistakes far more easily than hidden ones.
The opportunity hidden in distrust
Here’s the interesting part. If most people distrust AI, and most companies are implementing it badly, what happens to the company that implements it well?
It gains a competitive edge. Simple as that.
While your competitors deploy chatbots that annoy customers and voice agents that sound like cheap telemarketing, you can build systems that genuinely save time, improve the experience, and free your team to focus on what matters. The bar is low. Clearing it doesn’t require genius — it requires thinking things through.
How we do it at Studio SmartWork
Our way of avoiding all this is pretty simple: we don’t sell software, we don’t sell templates, and we don’t sell hype. We start by listening to the problem, design a concrete solution, build it with open-source tools so you’re not locked in, and get it running in under a week. Then we monitor it and improve it.
Real client results:
- Lead follow-up: from hours to under 60 seconds response time. Pipeline +22%.
- Inbox management: from 3 hours to 15 minutes a day. Escalations reduced by 90%.
- Sales proposals: from 1–2 days to 10 minutes. Capacity tripled.
These aren’t magic numbers. They’re the result of applying AI where it makes sense, with human oversight, transparency, and ongoing maintenance.
In short
People distrust AI because they’ve seen too many bad examples. But well-made AI is not the kind that makes headlines — it’s the kind that works quietly, frees up time, and doesn’t need defending because the results speak for themselves.
If you’re thinking about automating processes in your business, the question isn’t whether to ride the wave. The question is: will you do it well, or will you add to the distrust statistics? The difference between the two isn’t budget. It’s approach.
Start with the problem. Measure the wasted time. Design with care. Keep humans at the center. And above all, don’t believe the hype — not even ours. Ask for proof, ask for transparency, ask for results.