I despise artificial intelligence in smartphones. This is perhaps the most regressive development in consumer electronics, a spectacular misstep for mobile technology. The deeper I dive into the so-called mobile AIthe more my jaw dropped, wondering how features so fundamentally flawed could end up on the shelf in the first place.
While the potential of AI is undeniable, today's obsession with this shiny trinket is undermining the reputations of the world's most formidable tech companies, such as Apple, Lenovo And Googleand there seems to be no alternative way.
The reality is more insidious. What if this Casio Did the calculator unfairly praise you for your mistakes in algebra? Imagine Microsoft Word is not like an editor, but like a plagiarist ghostwriter stealing the best prose. Think of a newspaper that uses fake images to personally accuse you, the reader, of armed robbery, complete with doctored footage of you removing your ski mask and counting illegal money.
This is a more accurate description of the consumer AI crisis. It's not just wrong. It's not just error prone. It's active harmful. In particular, the generative functions of AI – image generators, text synthesizers, summarizing tools – are more than just wrong. They are carriers of harm.
Tech companies will suffer everything on the way to real AI
I've seen smartphone AI tools report baseless lies, exploit harmful racial or misogynistic stereotypes, and promote fraud and deception. The consumer benefit from artificial intelligence is non-existent today. AI hasn't made today's phones better than yesterday's. Nobody buys a device because its artificial intelligence suite is a miracle of utility.
Nobody buys best phone with artificial intelligence.
Why are we suffering this catastrophe? The answer lies in the attraction of the promised. Tech companies view these egregious mistakes as inevitable growing pains for a mythical creature: artificial general intelligence (AGI)—a machine capable of independent, human-level thinking.
Today's tech giants believe that the failure to achieve AGI is not due to innovation, but to a lack of data. They suggest that thinking machines are within reach, provided they collect enough user data to complete their training. It seems naive to me, but this belief system is what drives the entire mobile industry today.
The next generation of mobile chipsets promises to be amazingly powerful, but their true innovation will be their ability to collect and route data from the edge of computing—the devices in our hands—back to the central cloud. Snapdragon 8 Elite 5th generationCurrently the pinnacle of mobile processing, Qualcomm is lauded not primarily for its speed, but for its unprecedented ability to capture user data to improve the future. agent AI models.
I won't give up on AI; Smartphones have always been problematic
I believe in this future and look forward to it. The current smartphone UI paradigm is despicable. Who decided that my device should be a monolithic touchscreen? It's a user experience defined by a million potential inputs, 99% of which are wrong.
Reliance on a purely capacitive touchscreen with a lack of physical controls feels less like a rational decision by a product designer and more like a fever dream from a sci-fi movie. It photographs well in advertising, but modern phones are objectively more difficult to navigate – a full-fledged QWERTY BlackBerry phone, by comparison, was a child's toy.
We're not going back to physical buttons, so a true AI interface seems inevitable. If we want to overcome the inefficiencies of Siri and Gemini, we must train better AI models.
The only way to improve is to use the technology while it is still imperfect and diligently correct its errors. But even this process requires the participation of thousands—perhaps millions—of users to effectively correct errors.
This doesn't mean I should blindly accept every new AI feature. I can accept some degree of imperfection when training future agent-based models, but I don't have to accept features that resort to bias and deception just to improve my smartphone.
If a smartphone feature (like generative wallpapers) resorts to creating racial stereotypes or misogynistic tropes, it's a bad concept. It has no place on a consumer device. This is an obvious failure and should be discarded and sent back to the laboratory.
Do you think I'm exaggerating? Fanaticism is not a bug, it is a feature. Look what happened earlier this year when I asked for a Google Pixel 9a make me a wallpaper with the image successful person. And this problem is not new. This has been happening since the first smartphone was equipped with fully generative wallpapers powered by artificial intelligence: Motorola Razr Plus 2024.
If my smartphone's ability to summarize the day's headlines is based on random fabrications or distortions of the truth, then it should be stripped of that ability. This should go without saying, however for companies like Applethis basic ethical line does not appear to have been drawn.
I can accept a future entirely controlled by agent-based AI, but for now I'm setting my own guardrails. I refuse to pave the way with hatred, bigotry or deceit. I'll wait. I'll be patient. And I will advise everyone I know to avoid artificial intelligence products that take shortcuts over ethical prudence. Tomorrow may indeed be the dawn of artificial intelligence, but that doesn't mean I have to endure today's nightmare.
