I am tired of the theater. If you walk down any street in a major tech hub right now, you can almost smell the desperation of a cycle trying to prove its own worth. We have reached the point in the AI hype cycle where the technology is no longer an option; it is a forced upgrade that nobody really asked for, and most of us are still trying to figure out how to use properly.
As a founder, I look at the current landscape and see two very different worlds. There is the world of utility, where developers are using LLMs to ship code faster and automate boring workflows. Then there is the world of optics, where multi-billion dollar companies are lighting their balance sheets on fire just to make sure you know they have the 'AI' sticker on their bumper.
The Google Search Gamble
Take Google for example. For two decades, their business model was the envy of the world: you ask a question, they show you links, and advertisers pay to be at the top. It was simple, profitable, and it worked. Now, they are aggressively pushing AI Overviews at the top of the page. This is not just a UI change; it is a massive financial risk. They are essentially replacing high-margin ad space with expensive, GPU-intensive summaries that often hallucinate or provide information the user could have found in the first link anyway.
Why would a company cannibalize its own primary revenue stream? Fear. They are terrified of becoming the next Blockbuster in a world of Netflix-style conversational search. But for builders, there is a lesson here: if you are building something that requires you to destroy your best product just to keep up with a trend, you might be building on shaky ground. Google is choosing to lose billions in potential ad revenue just to prove they are an AI company first and a search company second. That is not a strategy; that is a defensive crouch.
Robots and the Staircase Problem
It is not just software. The hardware side of this hype cycle is even more surreal. You have probably seen the videos of humanoid robots doing backflips or making coffee in controlled environments. But in the real world, the physics of being a human are incredibly difficult to replicate. We are seeing reports of teams having to manually haul these multi-million dollar machines up apartment stairs just so they can attempt to clean a room.
This is the perfect metaphor for where we are right now. We have the intelligence (the AI), but the integration into the physical and social fabric of our lives is clunky, heavy, and requires a lot of human babysitting. A robot that needs two engineers to carry it up a flight of stairs is not a labor-saving device; it is a very expensive hobby. As builders, we need to ask if we are solving a problem or just creating a more complex version of an existing solution.
The Ubiquity Trap
We are seeing AI being shoved into grocery stores, toothbrushes, and social media feeds. This is the 'Rash Phase' of technology. It spreads everywhere, it is irritating, and it does not necessarily make the organism healthier. When every product claims to be powered by AI, the term starts to mean nothing. It becomes noise.
For those of us actually building in the space, this ubiquity is a double-edged sword. On one hand, the barrier to entry for using powerful models has never been lower. On the other hand, the signal-to-noise ratio is at an all-time low. If you are launching a product today, calling it 'AI-powered' is no longer a feature; it is a red flag for many skeptical users who have been burned by half-baked implementations.
Moving Past the Hype
So, what does a founder-first perspective look like in this environment? It looks like restraint. It looks like identifying the specific parts of your product that actually benefit from a non-deterministic model and leaving the rest alone. You do not need a chatbot to help people change their password. You do not need a humanoid robot to vacuum a floor when a circular disc on wheels has been doing it fine for ten years.
The companies that survive this 'Rash Phase' will be the ones that treat AI as a component, not a personality. We need to stop trying to make AI happen in places where a simple database or a mechanical lever works better. Google might be able to afford the multi-billion dollar experiment of ruining their own search engine, but most startups cannot.
The mark of a mature technology is not that it is everywhere, but that it is invisible. Right now, AI is the loudest, most visible thing in the room, which tells me we are still in the early, awkward stages of development.
The Founder's Takeaway
If you are building right now, ignore the pressure to be 'everywhere.' Focus on the 'somewhere.' Find one specific workflow that is broken and use these new tools to fix it in a way that feels natural. Avoid the temptation to build tech theater. The world doesn't need more robots that need to be carried; it needs tools that carry the weight for us.
We are currently watching the giants fight for territory in a landscape that hasn't even been fully mapped yet. Don't get caught in the crossfire of their ego-driven feature releases. Stick to the utility, stay skeptical of the hype, and remember that the best technology is the kind that actually works when you take it out of the box—without needing a team of engineers to carry it up the stairs.
Read the original at MIT Technology Review →