Saturday, November 23, 2024

I tried Google's suggested uses of Gemini Live – and one immediately stood out

Must read

Sabrina Ortiz/ZDNET

A human-like voice assistant that speaks to you sounds like something from a sci-fi movie. However, the technology is already here, with AI-powered voice assistants like Gemini Live easily accessible from your phone. So, how can these assistants help with your everyday life?

Although talking to someone is a cathartic experience, chatting with AI doesn’t exactly accomplish the same goal, as you know you are talking with a robot. As a result, despite being fascinated at how good Gemini Live is at understanding what I say, I often wondered about its helpfulness. 

Also: This absurdly simple trick turns off AI in your Google Search results

To help, Google has released a list of five ways Gemini Live can make users’ lives easier, and I tested each one. Below, you can find their list, ranked by the ones I found the most to least helpful, as well as a recap of my experiences. 

1. Creating a to-do list 

One of my favorite ways to use ChatGPT is to develop basic lists, such as what to buy at the grocery store and what to take on vacation. Typically, when I use this feature, I type my request into the chatbot. However, with Gemini Live or the Advanced Voice Mode, you only have to ask in conversation and have the assistant output your list. 

Also: The best AI chatbots of 2024: ChatGPT, Copilot, and worthy alternatives

The biggest pro with this approach is that, similarly to having a normal conversation with a human, you can stop and ask the AI to elaborate, add something else, restart, remove something, or tailor the list more to your liking. Depending on your situation, the bot will suggest things to add or other to-do lists. 

For example, my first prompt was, “Help me make a to-do list for Thanksgiving.” The bot’s response was a suggestion to start making a grocery list. Chatting like this is a more fluid interaction and takes less time as you are not glued to your laptop or keyboard. So, I agree with Google that this is a good use case and I think it’s the most practical one. 

2. Directed breathing exercises 

Although this suggested use case got the number two slot, it was a close runner-up for one. In the busy hustle and bustle of life, finding time to take a step back and relax or prioritize mindfulness is key. For that reason, most mornings I start my days with an affirmation session — including breathing exercises led by someone on YouTube. However, these experiences aren’t tailored to me and can get repetitive; that’s where Gemini Live can help. 

Also: Google’s Gemini Advanced gets a very useful ChatGPT feature – but how does it compare?

I asked the bot, “Can you lead me in affirmations for the morning for success?” Without skipping a beat, the AI generated a few affirmations I could dictate to get started. 

Then, to keep the session going, you can say something like, “Can you give me some more?” or ask it to give you a different type of affirmation. The same goes for asking for breathing exercises. The bot will generate a catered breathing routine. As someone who relies on guided mindfulness, this approach seems like a game-changer. 

However, that’s the good news — and the rest of the use cases go downhill from here. 

3. Rehearsing a presentation

Some things can objectively improve a presentation, including the pace at which you speak, your conciseness, and the engagement of your script. For that reason, Gemini Live can listen to your presentation and advise on how to improve. It can also be helpful to rehearse in the privacy of your own space without having to use another human to shake off the nerves.

Also: AI just gave us the Star Trek farewell we always wanted – watch it here

However, I found the experience a bit unnatural. I still think a human can give better advice as other factors that affect a presentation include posture, facial expressions, movement, and, most importantly, the presentation itself. If you have a terrible slide deck, not much can save you. If your goal is to improve speech, then this feature is fine — that’s about it, though.   

4. Roleplaying 

As one of the use cases, Google encourages users to leverage Gemini Live to have a mock difficult conversation. All a user has to do is explain the situation briefly to Gemini Live and start the conversation, receiving feedback along the way. 

For my case scenario, I said, “Can you help me role-play a situation in which I tell my best friend I don’t like her shoes?” The bot asked about her name and gave me a sample sentence I could use to tell her, but I don’t think the conversation would have landed well if I had used it. 

Also: How to use ChatGPT to digitize your handwritten notes for free

“Alright, here’s one way to handle this situation: Hey Melissa, those shoes are really something else! I’ve never seen a style quite like that before,” said Gemini Live. 

If my aim was passive aggressiveness, the bot nailed it. When I replied with, “You don’t think that’s mean?”, the AI answered: 

You’re right; it might sound a little sarcastic. How about this: “Mel, I noticed your new shoes. They’re really bold and eye-catching.” You could also ask her where she got them. This way, you acknowledge them without giving your opinion on whether you like them or not.

However, correcting the approach defeats the purpose of roleplaying. Usually, you roleplay with a friend when you don’t know what to say or do. I would prefer to use a human in this situation because the interaction is more natural and I ensure I get the best advice. 

5. Picking what to wear 

Gemini Live does not yet have multimodal abilities (where it can see your surroundings). For that reason, this use case finishes last. In my opinion, it is completely useless. Moreover, for the bot to help you pick out your wardrobe, you need to describe your outfit in detail, and many different factors need to be considered, such as texture, patterns, and fits, which can really only be described visually. 

How to access 

If any of these use cases stood out to you, you can access Gemini Live for free via the Gemini app on both iOS and Android. Once you’ve downloaded the app, sign in to your Google account, click the waveform icon, and start chatting. 

Latest article