Where is the AI gold? Nobody 10x’d Anything (Yet)!

(This post has originally been published on LinkedIn and has been cross-posted here for conservation and accessibility.)

🪏 Where is the AI gold? Nobody 10x’d Anything (Yet)!

There’s a gold rush happening in software development right now. My feed is full of developers sharing their “10x workflows” — new agents, new frameworks, new tools — all promising to multiply your output by some breathtaking factor. And just like in a real gold rush, the excitement is contagious. Nobody wants to be the one left standing at the river with an empty pan.

I use AI tools every day. Claude Code, Gemini and more are part of how my works. I’m not here to tell you AI is useless. It’s not. It’s genuinely exciting, and I happily plan to keep using it.

But I am here to ask one uncomfortable question: Where is the gold?

Not the vibes. Not the demos. Not the “I built a landing page in 20 minutes“ tweets (do we still call them that?). I mean actual, measurable results at a company level. Revenue. Output. Value creation. The stuff businesses run on.

Because so far, no company has credibly demonstrated that they 2x’d — let alone 10x’d — their output by replacing human effort with AI tooling. The productivity gains people cite are almost always self-reported, based on individual coding tasks, and disconnected from any business outcome a CFO would recognize.

The companies that are printing money from AI? OpenAI, Anthropic, Nvidia are great companies building genuinely impressive technology. But in gold rush terms, they’re the shovel sellers — not the diggers. Their success doesn’t prove that the rest of us will find gold. It proves that selling shovels works. And the fact that so many of us are afraid of being left behind certainly doesn’t hurt their bottom line.

Here’s the thing most of us won’t say out loud: most software developers are not ML engineers. Most aren’t even hobbyists in that space. And yet every new model release, every trending thread on X or LinkedIn creates this pressure to drop everything and retool your entire workflow. Again. That’s not just unreasonable — it’s unhealthy.

Here’s my principle (that applies well beyond AI) : You don’t want to be the first, biggest, or most complex user of any technology. You want to use what is proven to work — ideally by people far more incentivized and experienced in that domain than you are.

I’m not saying sit this out. I’m saying stop sprinting. Be curious, read up on what’s happening — but maybe don’t rebuild your workflow every week based on someone else’s X thread.

As with every gold rush, a few people will strike it rich. But historically, most everyone else eventually burns out. And unlike gold, burnout doesn’t have a market price that makes the suffering worthwhile.

Be careful out there.

As usual, I am happy to be proven wrong - there is no better way to learn! If your company has indeed 10x’d their value output, please let me know! If you’re still stuck at mostly individual wins that are hard to quantify - let me know too!

Photo by Drei Kubik on Unsplash