The Night I Almost Bought a Mac Mini I Didn’t Need

There I was, at 11pm on a Tuesday, with the Apple cart open in my browser. A Mac Mini M4 Pro configured. $1,599. Ready to close the purchase.

Why? Because in the previous weeks, I’d seen Andrej Karpathy talking about OpenClaw. I’d seen Simon Willison buy a Mac Mini “specifically to play with Claws.” Apple Store salespeople said these devices were “selling like hotcakes.” The feeling was clear: if I didn’t have a dedicated Mac Mini to run local agents, I’d be missing a revolution.

Luckily, that night, something made me close the tab before hitting “Buy.” Maybe it was tiredness. Maybe a question that sprouted on its own:

“Mauricio, what real problem are you trying to solve with this Mac Mini?”

The honest answer: none. I wanted to buy the machine to run a tool (OpenClaw) I still didn’t quite know how I’d use, to achieve a status (“be a relevant AI dev”) that was more performative than practical.

I almost spent $1,599 chasing hype before I had a problem that needed solving.

And I discovered, in the weeks that followed, that I was far from the only one.

The “Great Productivity Panic of 2026”

On April 5, 2026, Bloomberg Businessweek published a cover story that named what I and many others were feeling: “The Great Productivity Panic of 2026.” The diagnosis was precise:

“Vibe coding was supposed to be chill. But one year later, the vibes, as they say, are off.”

Bloomberg documented something counterintuitive: AI tools are working. Productivity gains are real. But instead of letting workers go home earlier, these tools are keeping them at their desks longer. A UC Berkeley study cited by Bloomberg followed a 200-person technology company and found: employees using AI tools increased the volume and variety of work — they didn’t reduce effort.

The ADP Research Today at Work 2026 report, with over 39,000 workers across 36 countries, arrived at a central paradox: daily AI users are 4 times more likely to say they’re “not as productive as they could be” compared to non-users.

The explanation came from Nela Richardson, ADP’s chief economist, and it hit me hard: “AI does exactly the things that we used to say make us feel productive.”

Reading emails. Organizing tasks. Responding to messages. Drafting reports. All of it now automated. And without those “little triumphs,” we feel strangely empty — even while delivering more. It’s a particular kind of existential fatigue no 2020 study would have predicted.

Motion vs. Progress

This is the distinction I took too long to make.

Motion is testing every new AI agent that launches. Subscribing to Cursor. Configuring Claude Code. Swapping Cursor for Windsurf. Swapping Windsurf for a new IDE that promised to “change everything.” Configuring OpenClaw. Trying NanoClaw. Posting about it on Twitter. Reading threads on “context engineering” at 2am.

Progress is choosing a direction. Facing the boring, complicated parts of a project. Shipping something with real users. And doing that long enough to see compound results.

The difference looks subtle. The results are abyssal.

And Silicon Valley’s machine in 2026 runs on your fear of stopping. As one developer posted on X: “I scrolled my Twitter feed and counted: in one day, I saw 23 separate posts saying ‘you need to try this AI tool.’ Twenty-three. From accounts I follow. That’s not organic excitement. It’s an attention economy extracting engagement from developer FOMO.”

Google’s DORA Survey 2025 of nearly 5,000 professionals found that 90% used AI at work and over 80% reported productivity gains. The gains are real. But between the study and your midnight Twitter feed, there’s a huge gap. And that gap is where the anxiety lives.

Anxiety as a Business Model

Here’s the line that changed how I think about all of this:

“Anxiety is the product.”

When you subscribe to an AI newsletter, the metric that matters to the publisher isn’t your well-being — it’s your open rate. When you follow an AI influencer, the algorithm rewards content that generates engagement — and nothing generates more engagement than urgency.

The result is an entire ecosystem optimized to make you feel like you’re falling behind. Endless threads. Weekly releases positioned as “revolutions.” Headlines like “If you’re not using X, you’re losing out.”

And the consequences are measurable. A Bloomberg piece from March 26, 2026, documented something important: employers are using both productivity gains and AI threats to push workers harder — instead of redesigning work to be sustainable. More output, same people, no additional support, no adjusted expectations about what a reasonable workload looks like now.

The “AI Shame” phenomenon completes the picture: 62% of Gen Z workers hide that they use AI. 55% pretend to understand tools they haven’t mastered. Another sign that this ecosystem isn’t producing confidence — it’s producing performance.

Tool Tourism: The Anti-Pattern I Fell Into

An analyst described it well: “tool tourism.” Constantly switching stacks to hide a weak positioning. If you’re always “testing the next big launch,” you never master any tool enough to ship something exceptional with it.

And the cost isn’t just time. AI tools frequently ship in “permanent beta.” A developer described it: “They launch with a flashy demo. The core feature works. Then you try to integrate it into your workflow and discover that half the documentation is wrong, the API changes every two weeks, and the feature you signed up for gets deprecated in favor of a ‘new direction’ after three months.”

I saw this pattern happen with at least four tools where I invested real time in the past year. Switching cost is invisible on the surface — but brutal in aggregate.

The Value of Being Somewhere Real

Things that really last in technology generally come from someone who decided to follow a single direction for longer than seemed sensible.

It’s not about ignoring AI. It’s about using it with intention. After the Mac Mini night, I adopted some rules that have been working:

I don’t try to be everywhere. I don’t need to be an expert on the 23 models topping the leaderboard today. I need to know well the 2 or 3 that solve the problems I actually have.

Curious, but strategic. Curiosity is fuel. But it needs to be pleasant, not distressing. If I open an AI newsletter and close it feeling worse than before I opened it, that newsletter isn’t serving me — it’s feeding off me.

Solve problems, not feelings. This is the most important one. Before downloading the next autonomous agent, configuring the next tool, or buying the next hardware, I ask: “What real pain am I trying to cure with this?”

If the answer is “I don’t know, but everyone’s using it,” that’s a clear red flag. FOMO feelings don’t justify $1,599 decisions.

What I Did Instead of Buying the Mac Mini

That night, I closed the tab. The next morning, I opened a blank document and wrote three columns: “Problems I have,” “Tools I already use,” “Real gaps.”

It was an honest exercise. And to my surprise, most of the tools I already subscribed to covered my real problems. What was missing wasn’t a new tool — it was discipline to use the ones I already had well.

The $1,599 was redirected: half to a specific course in an area I wanted to go deeper, the other half stayed in savings. Six months later, I can say without doubt it was the right decision.

Conclusion: Thinking Is Your Edge

In a world moving so fast that “thinking seems to be falling behind,” the biggest competitive advantage for a professional in 2026 is the ability to stop, filter the noise, and focus on what’s real.

AI is powerful. The productivity gains are real. But the way you adopt AI — with FOMO or with intention — makes all the difference between gaining time and burning money chasing a feeling that was sold to you.

You don’t need every AI. You need the right AI for the right problem.

And the first step to finding that AI is to stop looking at what others are using and start looking at what you’re actually trying to build.

And you? Have you ever bought hardware or subscribed to an AI tool just for the “vibe” and later realized you had no use for it?

I almost did. Luckily, I stopped in time. But the shame of almost having done it is what keeps me skeptical today — and honest about what I actually need.

Share your story:

AI solves many problems. But feelings aren’t problems — and confusing the two costs a lot.

Read Also