Science fiction has always made it seem like talking to computers is the pinnacle of user interaction. Shows from Knight Rider toStar Trekhave featured characters having seamless conversations with their tech, allowing them to do everything from calling their self-driving cars to making cups of Earl Grey (hot).
As with so many things these days, what was once the realm of science fiction is now reality.Voice assistantsthat allow us to talk to our phones and computers have been around for some time, and recent developments in AI have seen these voice assistants become even smarter.
Just last week, Google launchedGemini Live, a conversational voice assistant with almost instant response that’s available right now onAndroid phones. OpenAI is in the process of rolling out its own updatedChatGPT Voicemode. With major players such as Apple, Google, and OpenAI going big on voice capabilities, it throws up a genuine question: Do we actually want to talk to our phones and other devices?
10 Gemini Live features I can’t wait to try
Google’s AI sounds more human-like, but what, exactly, is Gemini Live capable of?
Gemini Live is the latest entry into the world of conversational voice assistants
Google Assistant has been replaced with a much smarter version
Google first introduced updated capabilities for Gemini atGoogle I/O 2024back in May. During the recentMade By Googleevent that saw the launch of the newPixel 9phones, we finally got to see Gemini Live in action, alongside an announcement that the new features were rolling out the same day for Android phones.
Some of the best features are not yet ready, however, such as the ability for Gemini to integrate with apps such as Keep, Tasks, and Calendar. As yet, Gemini Live is also unable to pull information from a live video feed, which was arguably the most impressive feature showcased back in May.
The live demonstration was still reasonably impressive, however. We were shown how you can choose from 10 different voices forGoogle Gemini(none of which sound remotely like Scarlett Johansson) and then Gemini Live was used to brainstorm some ideas for something fun and educational you could do with your niece and nephew. Gemini Live offered suggestions such as creating a homemade volcano and making your own invisible ink.
It’s highly unlikely that anyone would truly love talking to an AI assistant, even one that’s much more natural than before.
Jenny Blackburn, who is in charge of Gemini experiences at Google, was able to go back and forth with Gemini Live to get more details on the experiment, whether or not it would be messy, a creative name idea for the experiment, and other useful information. The conversation sounded natural without the long pauses that have plagued voice assistants up to now, and she was able to interrupt if necessary.
At the end of the demonstration, Blackburn gave her thoughts. “This experience is so cool. I love talking to Gemini in a free-flowing conversation that can go in any direction.” Her words didn’t entirely ring true, however. This isn’t surprising, because it’s highly unlikely that anyone would truly love talking to an AI assistant, even one that’s much more natural than before.
5 cool things Google’s Gemini AI can do on your Pixel 9
The new Google Pixel 9 phones have some exclusive AI features.
ChatGPT Voice and Siri are both about to get smarter, too
Updates to ChatGPT Voice and Siri are just around the corner
Gemini Live is clearly trying to get a jump on the latest update to ChatGPT Voice, which was announced back in May but has yet to roll out to more than a handful of users in preview. The new features ofChatGPT Voiceare remarkably similar to Gemini Live; almost instant response, the option to interrupt, and the ability to gather information from a live video feed. However, Gemini Live has managed to get those first two features into the wild, with ChatGPT Voice still yet to receive those upgrades. It seems like it will be some time before either model is able to make use of live video.
We shouldn’t forget the voice assistant that started it all, either. Siri may not be gaining quite the same capabilities as Google Gemini or ChatGPT Voice just yet, but iniOS 18, there will be significant improvements that make talking to Siri a much more conversational process. For starters, the assistant will finally be able to remember the context of previous parts of the conversation, so you won’t have to repeat key information with every prompt.
How I upgraded Siri with ChatGPT to get smarter AI responses on my iPhone
I can still talk to Siri, but now I get better answers generated by ChatGPT. It’s the best of both worlds.
There are plenty of reasons why we might not want to talk to our devices
Voice interaction comes with a whole host of issues
With companies going big on AI voice assistants, the question is, do we actually want to talk to our phones and computers? We’ve had the ability to do so for a long time.Sirihas been around since 2011, Alexa has been part ofEcho speakerssince 2014, and I was able to build a simple version of J.A.R.V.I.S. from an Xbox Kinect sensor and a laptop back in 2015. But to paraphrase Jeff Goldblum, tech companies were so preoccupied with whether or not theycouldget us to talk to our devices, they didn’t stop to think if theyshould.
I have Echo devices in my home. I have Siri on my wrist. I have ChatGPT Voice. I have a voice assistant within myHome Assistantsetup that I can use to control my smart home beyond anything that Siri or Alexa can do.
Here’s the issue: I just hardly ever use them. Talking to my phone or my computer doesn’t make me feel like Michael Knight or Jean-Luc Picard, even though I’vechanged the wake wordon my Echo devices to “Computer.” It makes me feel like a self-conscious idiot who’s talking out loud when he shouldn’t be.
Tech companies were so preoccupied with whether or not theycouldget us to talk to our devices, they didn’t stop to think if theyshould.
I don’t think I’m alone; several Pocket-lint colleagues have also confirmed that they hate talking to their devices. Anecdote isn’t evidence, but there’s something else that might be. Even before the smartphone arrived, we had two choices if we wanted to contact some with our phones. We could call them, or we could message them. Guess which one we did most?SMS messagingwas only intended as a secondary feature, but by 2007 had overtaken calling as the most popular way to communicate on a phone. Messaging apps such as WhatsApp eventually took over from SMS, but the underlying principle remains the same. When given the choice,we would rather type out a message than speak it out loud.
It’s embarrassing enough talking to a machine when you’re on your own, but what happens when you’re out in public? I honestly can’t think of a single occasion where I’ve used a voice assistant out loud in public when other people were around, and I can’t think of a single occasion when I’ve seen anyone else do it, either. Holding a full-on conversation with ChatGPT or Gemini when other people are around? The idea is just mortifying.
Using voice has its place in the right circumstances
There are times when voice is definitely more useful
That’s not to say that powerful voice assistants such as ChatGPT Voice and Gemini Live don’t have a time and place. There are some instances when they’re infinitely more useful, such as when using your hands isn’t really an option.
Being able toask Alexato start a timer for cooking the chicken using my voice is infinitely preferable to pulling out my phone with the same hands that have just been cutting up said chicken. Being able to use a voice assistant when you’re driving is also invaluable, allowing you to do things that otherwise wouldn’t be possible when you’re behind the wheel, such as opening your perfectly curated road trip playlist that you forgot to start playing before you set off.
I’ve also seen some posts on Reddit in which users talk about what they use AI chatbots for, and an answer that surprised me was that some people use chatbots simply as someone to talk to. A realistic voice chatbot that feels like talking to a real person could genuinely be a useful support tool for people with social anxiety or who are simply lonely.
Ultimately, voice assistants can be useful in the right circumstances.
I’ve also been using ChatGPT Voice to practice speaking Italian; it’s a useful tool as you can prompt the AI to gently correct you whenever you say something that isn’t quite correct, and you don’t have to worry about sounding like a fool in front of a real Italian speaker. It offers me the chance to get a lot more speaking practice than I do withDuolingo.
Ultimately, voice assistants can be useful in the right circumstances. However, it does feel like major companies seem to think that we crave the ability to interact via voice more than maybe we actually do. When those companies are asking us to pay for the privilege, which is currently the case with both Gemini Live and ChatGPT Voice, it feels like they might be flogging a dead horse. The good news is that, according to science fiction, we’ll eventually be able to control computers with our minds, in sweet silence. Bring it on.
