Sometimes, the Gemini Assistant’s mere thought of opening another application on my phone and typing in some text feels like an overwhelming emotional burden. Whether it’s gathering information from multiple confirmation emails for an Airbnb reservation or figuring out the optimal time to leave a coffee shop to catch a bus, these tasks may not be difficult per se, but they involve navigating through various apps or switching between screens. In such moments, I find myself questioning the necessity of sending that email right away. Perhaps I can just improvise and hope for the best when it comes to the bus schedule.
These are the responsibilities that I wish AI could handle on my behalf. While Google’s new Gemini assistant and other AI technologies are not quite capable of fulfilling these tasks yet, they provide a glimpse into the potential future of AI, especially for those deeply integrated into Google’s suite of services.
Gemini Assistant, previously known as Bard
It is an AI chatbot developed by Google. It is available for download on the Google Play Store, although it is essentially a component of the Google app that is likely already installed on your Android phone. Once installed, Gemini can be used as a replacement for the standard Google Assistant, and can be invoked in the same manner as its predecessor. However, unlike the old Assistant which primarily handled tasks such as setting timers and providing weather updates, Gemini is capable of answering intricate queries, offering suggestions, and even reading your emails, with your permission.
The aforementioned final part holds significant importance. ChatGPT surpasses Gemini in terms of conversational skills, however, Gemini’s capability to integrate with Gmail, Google Maps, and Google Docs is what truly captivates. I requested Gemini to summarize the specifics of the Airbnb reservation, and it successfully extracted information from two separate emails and presented it in a concise bullet point format. Subsequently, I instructed it to compose an email to my friends, incorporating all the necessary details. Usually, when I rely on AI to write an email or text, the outcomes are too embarrassing to share with anyone. To my astonishment, this particular result was satisfactory.
I must admit, it may not seem like a significant achievement, but this is the first instance where AI has truly amazed me as a useful tool. Perhaps I lack creativity, but I find myself growing tired of ChatGPT rather swiftly; there are only so many occasions where I can brainstorm business strategies for a vintage arcade or contemplate vacation ideas. What I truly desire is assistance in managing the overwhelming amount of digital clutter that constantly engulfs my daily existence.
Regrettably, Gemini has not reached that level yet. In my opinion, it surpasses the regular Assistant by a significant margin in terms of usefulness, and there are many aspects that I find appealing about it presently. One notable feature is the small G icon located at the bottom of each response, which enables you to search Google and fact-check Gemini’s information. Given that AI occasionally tends to provide inaccurate details, this feature holds considerable importance.
Furthermore, I value the fact that Gemini provides a designated space for me to store any information it generates, such as recipes or packing lists. Gemini allows for direct export of answers to Google Docs or Gmail. On the other hand, when I receive similar outputs from ChatGPT, they seem to be floating aimlessly until I manually copy and paste them elsewhere. Although they are saved in your history, I believe you understand what I mean.
Gemini lacks the ability to grasp context effectively. I informed it about my intention to ride my bicycle to the adjacent neighborhood and requested recommendations for activities upon arrival. To my surprise, it generated an extensive list of suggestions, including scuba diving, attending live theater, and gambling at the casino. It is worth mentioning that while these activities are indeed available in Burien, Washington, they are not suitable for impromptu daytime excursions. Furthermore, these are not the kind of places one would casually visit on a bicycle ride.
In situations such as these, Gemini functions as a helpful “let me Google Maps that for you” tool. On the contrary, ChatGPT’s voice chat recommended that I explore a few parks or a coffee shop. It even inquired if I had any other plans for my outing, and when I mentioned wanting to purchase a book, it kindly suggested a specific bookstore. How considerate! However, later on, Gemini’s integration with Google Maps proved to be more advantageous: when I asked for a bus route back home, Gemini provided me with the correct route, whereas ChatGPT would have had me waiting for a bus that only arrives every 30 minutes.
Interacting with Gemini gives me the impression of conversing with a page of search results from Google. When I initiate the conversation by saying, “Hey Google” and pose a question, it responds audibly. Otherwise, I am left with reading the text. Moreover, the amount of text can be overwhelming, indicating that this assistant could benefit from an editor. I am also taken aback by the fact that Gemini lacks the capability to access my calendar, unlike Gmail and Docs which have extensions for it. Consequently, if I wish to add an event to my calendar, I am compelled to switch back to the regular Assistant. In such instances, I would rather create the calendar event myself.
Ten years ago, smart assistants such as Siri and Alexa were presented as the future of device interaction. However, their advancements have come to a halt within the industry. Currently, Gemini serves as an optional assistant. Nevertheless, it is conceivable that Gemini could eventually supplant Google Assistant as the default choice, particularly considering Google’s gradual reduction of Assistant features in recent times. Perhaps we have simply reached the boundaries of what non-AI voice assistants can consistently accomplish.
To me, replacing Gemini with the typical Google Assistant seems like a risk with minimal consequences. I have already taught myself to refrain from using voice assistants as they often fail to perform the tasks I require. While Gemini can still handle basic functions like setting timers and providing weather updates, it would be beneficial to rely on a more intelligent assistant for my daily activities. Hopefully, it can acquire some valuable insights from ChatGPT as it progresses. However, for now, I must attend my scuba diving lesson.