Last week both OpenAI and Google announced some big announcements with their AI products. Underpinning some of the biggest announcements was how users can interact with AI using voice and showing AI things in the real world. The vision is to show how tech can become a helpful companion through voice conversations. However, while voice assistants have been around for a while, their usage has been slow to develop. There is no denying the possibilities abound for this technology, but to date, there has yet to be scalable adoption.
There are some technical barriers to entry with AI, and it could be argued that the limitations of text-based AI will perpetually limit the capability. While AI can provide more and more functionality as we move towards more multi-modality, it still sits within a device with a restrictive interface. This is why several innovations in the tech space over the last few months have sought to bring AI beyond the device, making it more connected to the real world.
This started with the likes of Humane’s AI Pin and Rabbit’s R1 platform. Both attempted to move away from a screen for interaction and focus more on voice as a connection to AI. With OpenAI and Google’s latest announcements this week, the shift towards voice-centric interaction is ready to take another step in its evolution.
Both OpenAI and Google showcased how their new improvements can allow AI to ‘see’ the real world through a camera (or glasses in Google’s case). This can unlock several new possibilities, making AI more connected to challenges beyond a text prompt and start to answer more of what people are seeking AI to resolve. From communicating with people in numerous languages to helping people find their glasses, these AI assistants have a lot they can offer. There is, however, still a lingering question that could prevent their success: do people actually want to talk to AI?
Every major tech company has seemingly attempted to release a voice assistant in the past. The likes of Apple’s Siri, Amazon’s Alexa, Google’s Assistant, Samsung’s Bixby, and Microsoft’s Cortana have all come along to try and create a new way of interacting and connecting with people. Unfortunately, all have struggled to reshape how people leverage voice. Through a combination of a lack of robust information, misunderstanding of voice queries, and sometimes comical results, voice AI has yet to build much trust in its user base.
Historically Google has been perhaps the best positioned to harness voice to connect to information, as it sits on the richest data available. Google rolled out voice search years ago, and while the uptake is encouraging, the reality is that it has yet to reshape how we engage with search.
According to Synup, 27% of searches in the Google app are now conducted using voice. This is an impressive number, especially when factoring in the volume of Google searches and that mobile searches represent 63% of total Google searches. However, according to The Manifest, this is a sign of the decline in voice usage across search. In their research, The Manifest discovered that the volume of voice searches declined from 2018 to 2021. According to the research, 2018 found 53% of respondents stated using voice search more than once a week. By comparison, that number dropped to just 18% in 2021.
The smart speaker market is also an indication of tepid appetite for voice assistants. Amazon has dominated the market for years and offers a low entry point into the voice assistant market, with the cheapest Echo device costing less than $40 USD. However, according to Amazon, only 25% of American households own at least one Alexa device. This does represent 72 million users, so is not exactly a small audience. However, it represents less than half of Amazon’s Prime subscribers, with over 100 million Prime users in the US not accessing the Amazon’s voice assistant.
The cause of this slow adoption rate could be driven by the fact that, while the technology is relatively mature, the capabilities are still limited. All the existing tools are predominantly used as another way to quickly get an answer to a question or perform a task when a physical device is out of reach.
There is no denying the benefits that voice assistants offer. From removing language barriers to helping solve challenges in the real world, and a simpler way to access information quickly and easily are all great benefits. There are signs that voice to AI has the potential to reshape how we access information. This is why both Google and OpenAI have illustrated this evolution in both their latest announcements.
There have been several showcases that highlight the power of speech with AI, but audience appetite has been limited to date. As the proliferation of voice assistance grows in access, usage, and capability, there is little doubt that adoption can grow, and grow rapidly. But there has yet to be a fully realised connection with consumers. Will Google and OpenAI’s advancements be enough to finally convince more people to use voice with AI?