There is a pattern we have watched play out across thirty years of technology cycles. Every wave arrives wearing the same costume: extraordinary claims, a handful of real results, a lot of borrowed credibility, and a deadline. Act now or get left behind. The deadline is never real. The pattern always is.
Right now, five things are happening simultaneously in the tech world, and most founders are treating them as separate events. They are not. They are the same event, wearing five different shirts.
The AI consciousness debate is the clearest tell. MIT Technology Review points out that the entire "rogue AI" rhetoric is a trap: while prominent lab founders push for regulation of supposedly superhuman systems, the actual question of what these tools can do for your business gets buried under science fiction. We are spending enormous cultural energy on whether AI agents are "awake" instead of asking whether they can reliably handle your customer intake form. One of those questions helps you. The other sells books and congressional testimony.
Meanwhile, Tesla is at it again. In a briefing with JPMorgan at its Fremont factory, the company laid out a timeline that included "step-change" self-driving and humanoid robots for sale by 2027. These are not new promises. They are the same promises, re-dated. The journalists covering this are careful to note that Tesla has made and missed these exact claims before. And yet the market responds. The coverage runs. The cycle continues.
At the infrastructure layer, Qualcomm's acquisition of Modular gives the chip maker a software layer that decides where AI workloads actually run. This is the unglamorous, consequential kind of news that gets half the coverage of the robot story. Whoever controls the routing layer controls the economics. That has always been true. It is true now. But it is harder to make a keynote about than a humanoid on a factory floor.
And somewhere in the design community, a quiet post about SMIL animation timing charts is making the rounds. An old, overlooked spec. Works without JavaScript. Animates SVGs inside plain image tags. Nobody announced it. Nobody made a conference out of it. It just works. Has for years. Most people building on the web never knew.
That last one is the signal. While the industry argues about consciousness and humanoid robots, the tools that actually solve problems have been sitting there, unannounced, unmonetized, unglamorous. The pattern holds across every domain we have ever worked in. The loudest technology is rarely the most useful one. The most useful one usually has no keynote.
Here is what Epictetus understood, and what Marcus Aurelius carried into every crisis of his reign: you do not control what the world throws at you. You control what you pick up and carry. Marcus ruled during plagues and wars and barbarian incursions. He did not spend time debating whether the Rhine would flood again. He prepared for what he could prepare for and acted on what was real in front of him. The Stoic practice was not detachment. It was filtration. What is actually true here? What is noise?
That is the exact skill you need in a tech cycle. And almost nobody is practicing it.
We see this constantly with founders we talk to. Smart operators, real revenue, products that work, hitting a wall somewhere between $1M and $5M, and suddenly convinced they need to rebuild everything around AI because that is what the conference circuit is telling them. They are treating the noise as a requirement. They are about to spend six months and serious money on a "step-change" that may get re-announced next year with a later delivery date.
The question you should ask before any technology decision is not "is this real?" Most of these things are partially real. The question is "real for whom, and by when?" Self-driving cars are real. They are not real for a two-person music label trying to automate their royalty reconciliation in 2025. AI agents are real. They are not magic. They are a conditional tool that requires real engineering work to deploy usefully, not a product you buy and install and walk away from. The routing layer that Qualcomm just acquired matters enormously at scale. It may matter nothing to you for three years.
Hype has a clock. Every wave that has ever swept through the industry, from the browser wars through the app gold rush through crypto through the current AI supercycle, has operated on roughly the same timeline. Enormous narrative pressure for two to four years. A shakeout. A handful of durable tools that survive. Most of the noise is gone by year five, and the founders who waited for the signal instead of chasing the sound are the ones still standing.
The founders who lose are the ones who mistake urgency for importance. Tesla has been selling urgency about autonomy since 2016. The consciousness debate creates urgency about AI regulation. The chip acquisition creates urgency about who controls the AI supply chain. All of it is real urgency, for someone. The question is whether it is your urgency, in your timeline, with your resources.
Here is how we think about it for ourselves and for the operators we work with. Separate the infrastructure decisions from the narrative decisions. Infrastructure moves slowly and matters for years. Narrative moves fast and matters for weeks. The SMIL spec that nobody announced has been sitting there quietly doing real work. That is infrastructure. The Tesla robot announcement is narrative. Buy the infrastructure when it matures. Ignore the narrative until it does.
The second filter is concrete impact in twelve months. Not "this will change everything." What will it change in your business, measurably, by next June? If you cannot answer that question in two sentences, the technology is not ready for you, whatever the conference speakers say about it.
We have built and shipped products across every one of these waves. We have seen founders pour money into Web3 infrastructure in 2021 and rebuild it in 2023. We have seen the same founder spend eighteen months on a custom AI pipeline when a $29/month API call would have done the job. We have also seen founders who refused to touch any of it and woke up three years later having missed the tooling that would have saved them a year of engineering time.
The answer is not skepticism. The answer is calibration. Marcus Aurelius did not refuse to plan for war because war was uncertain. He planned carefully, from what was real, and kept his head when everyone else was either panicking or celebrating too early.
Set your clock. The hype cycle runs whether you watch it or not. The founders who know what time it is are the ones who get to choose when to move.