Control: Test, implement, or discard

They say Mario Andretti came out with that line:

“If everything seems under control, you’re just not going fast enough.”

Going fast demands a degree of chaos. But there’s a difference between pushing the limit and having no brakes or direction at all.

In the design world we’re living through exactly that — and it’s not healthy: constant anxiety. Drowning in promises, gurus, and clickbait terrorists preaching either eternal life or the flames of hell…

Every week something comes along that “changes everything.” Every day, someone announces that this time it’s real, that now it’s here, that productivity is infinite. And you’re thinking: did I miss something?

Too much swerving sends you into a spin. Too many spins leave you last on the Mario Kart track, waiting for the star to save the race, always with your tongue hanging out.

What I needed was a steering wheel. Or a rudder. Something closer to riding a Tron bike: where turning doesn’t throw you off the curve. Well… or the straight, in this case.

The problem isn’t speed

Technology should be at my service. Not the other way around.

And sure, to regain direction, it makes sense to lean on AI. LLMs are a probabilistic stream of non-deterministic information. Without clear rules, they dump everything on you.

With AI, closing taps works better for me than opening paths. Less “What else can it do?” More “What don’t I want it to do?”

What I did

I built a GPT dressed up as a framework that doesn’t look for new things. It evaluates. That’s it.

When a tool shows up, I run it through a decision filter:

• Does it reinforce my positioning?
• Does it help me decide better?
• Does it leave something reusable?
• What’s the opportunity cost?
• Can I ignore it for three months without anything happening?

If the answer to the last one is yes, it’s probably not urgent.

How it works

• Markdown for thinking.
• YAML and JSON for structure (because AIs like explicit formats).
• A scoring system from 1 to 3.
• 14-day experiments.
• Clear rules: integrate, sandbox, or discard.

Nothing heroic. Nothing revolutionary. But with brakes.

Result

Three weeks using it, and while I don’t have scientific metrics, I have something better:

1. I’ve ignored more things without guilt.
2. I’ve made more decisions with judgment.
3. I’ve stopped running races I didn’t want to run.

If it’s useful to you, it’s on GitHub and explained here: signal.3lastic.com.

Copy it. Break it. Adapt it.

Going fast is fine. Knowing where you’re headed is better. And good luck managing your FOMO.