In the midst of the AI explosion, voice assistants are still halfway there. Apple continues to struggle with the implementation of artificial intelligence. The company confirmed at the end of the week that Siri’s advanced functions would take “longer than expected“ to arrive, without giving a specific date but saying that we will not have any news until 2026.
The company is late to the game, but it is not the only one facing problems with its smart assistant. The great alternative to Siri is Gemini, a solution that most Android manufacturers are starting to implement in collaboration with Google and which is still very, very green.
Don’t expect the new Siri anytime soon. One of the reasons for Apple Intelligence was the new Siri. Native integration with ChatGPT, natural language understanding, analysis of the content of our phone to get to know us in detail… We were recently able to test the beta version of Apple Intelligence and the conclusion was clear: everything was either half-built or not there at all.
Weeks later, Apple confirmed that the smarter version of Siri “will take longer than expected.” Its artificial intelligence is still in beta phase and all the new features were expected to arrive this spring. That will not be the case.
Apple decided not to jump on the AI bandwagon when its main rivals were at a point of relative maturity, and these delays have led it to its current situation.
Its rival is rubbing its hands. Meanwhile, the response in Android is clear. This operating system is owned by Google, and Google has Gemini as an assistant enhanced with artificial intelligence. Thus, most phones (OPPO, Samsung, Xiaomi, etc.) sold in Europe come from the hand of Gemini.
This is the agent that replaces the classic Google Assistant (“OK, Google”) that we have been using on our phones for so many years, with the main difference being Google’s response to tools like ChatGPT.
Not all that glitters is gold. Gemini has improved a lot since February 2024. Gemini Live is now completely free, has no problem performing simple actions (alarms, searches, etc.), but it is still a long way from being a natural assistant.
One of its main problems is precisely that the distinction between Gemini and Gemini Live dilutes the very use we want to give it as an assistant. If, for example, I ask Gemini what I can do today, it will give me a particularly long answer. If I want it to stop talking (in addition, Gemini’s tone is quite robotic and unnatural) I cannot do so comfortably, since the only mode that allows interruptions is Gemini Live.
In other words, in a standalone app (like Gemini or ChatGPT) this distinction between conversational modes makes sense. In a fast, native assistant, everything should be available in the most accessible way possible. And no, if you tell Gemini if you can talk using Gemini Live, it doesn’t activate this mode, it just starts talking non-stop about what this mode is.
Gemini also lacks access to native apps (it only works through extensions, and there are very few of them at the moment). It can’t even make simple adjustments like turning the phone’s brightness up/down, and the same goes for the volume. It can’t even change basic system settings if we ask it to.
There are no other rivals in sight (yet). The only Android manufacturer that was betting on a conversational assistant was Samsung with Bixby. This assistant is still alive in One UI 7, but it is so secondary that Samsung itself preinstalls Gemini on its phones and its extensions are key to the operation of Galaxy AI.
In China, major manufacturers are starting to integrate DeepSeek as native AI, but for now, there is no advanced voice mode or native integration. Honor wants to change everything with its AI agent, one capable of making all kinds of requests, including the most important ones, those for native settings.