Somewhere around 2023, every product on Earth got a memo: add AI or die. Toothbrushes started "analyzing your brushing patterns." Blenders got "smart." There is, we are told, a refrigerator that uses machine learning to suggest recipes based on what's inside it, which is a complicated way of saying it looked at an egg and recommended an omelet.
We'd like to say we're above this. We are not. We built 22 AI tools, one of which only knows how to make toast. The difference, and we'll die on this hill, is that we're honest about the fact that none of it needed to exist. Everyone else is pretending their AI-powered juicer is a breakthrough. We are simply guys that looked at the hype cycle and asked, "what if we just... kept going, past the point of usefulness, on purpose?"
The actual reason, probably
Nobody adds AI to a toothbrush because the toothbrush was missing something. They add it because "AI-powered" tests well in a pitch deck. Investors hear "machine learning" and their pupils dilate the way ours do at a clearance sale. It's not a product decision. It's a Mad Libs for funding rounds: insert the phrase "AI-powered" into any noun and watch a valuation appear.
This is, incidentally, also how we got funded. We are not proud of this. We are, however, financially stable, which in this economy counts as a personality trait.
So what should actually have AI in it?
Genuinely: not much of what currently claims to. Most "AI features" bolted onto existing products are a chatbot wedged into a settings menu, answering questions the help page already answered, slightly worse, with more confidence. We know this because several of our own tools are, structurally, that exact chatbot — we just stopped pretending it was a feature and started billing it as the whole product.
If your AI feature's main benefit is a shorter investor slide, it is not a feature. It is a slide.
A fork does not need AI. A doorbell does not need AI. Your toothbrush was fine. It was always fine. The problem it did not have has not been solved by giving it an app.
Our proposal
We're not asking anyone to stop building AI products. We are, after all, a bunch of guys who built ORVOI entirely out of them, in the same way a demolition derby is built entirely out of cars. We're asking for one thing: if your AI doesn't do anything useful, just say so. Put it on the box. "This blender has AI. It does not need it. We added it anyway." That's a company we could respect. That's a company we *are*.
Until then, we'll be here, proudly regressing the entire field of artificial intelligence one toast-only recipe generator at a time.