Meet your AI partner: Google I/O likely to unveil groundbreaking Gemini advancements

Gemini is capable of far more than what chat or voice interfaces presently allows
An undated image of Google I/O event 2024 poster. — Google I/O
An undated image of Google I/O event 2024 poster. — Google I/O

Alphabet-owned Google I/O 2024 live will take place on Tuesday, and while the annual developer event will feature new improvements for mobile, home, and wearable devices, AI will take centre stage.

Google Gemini is the search engine giant's family of artificial intelligence models that are gradually taking centre stage in everything Google creates, from replacing Assistant on Android to powering analysis in search results.

Moreover, at Google I/O event, people may expect to see a new version of Gemini, increased integration across several products, and multimodal features added to the Gemini chatbot, allowing it to accept voice, code, music, and video for the first time. 

Citing rumours, Toms Guide reported that Gemini may also embrace some of the more popular capabilities of rival OpenAI ChatGPT, such as permanent memory across all chats.

However, with OpenAI releasing its new GPT-4.0 model with built-in voice assistant and vision functions, Google will have to play catch-up.

Read more: Google I/O 2024 — Count down begins

What can you anticipate from the Gemini models

The Gemini moniker refers to the underlying huge language models, the Assistant replacement for Android, the chatbot, and the AI autocomplete in the workplace.

Gemini comes in three different variations, which adds to the confusion.

  1. The first is Nano, which works on phones and tiny devices.
  2. The second is Pro, which runs in the cloud and powers the Assistant; The free Gemini chatbot.
  3. On paper, the Ultra model is the most powerful, and it powers the Gemini Advanced, which costs $20 a month.

Google debuted the Gemini Pro 1.5 earlier this year. This was a significant enhancement over the previous version of Gemini since it included improved comprehension, music and video input, and a large million-token context window.

Gemini Pro 1.5 is solely available to developers and researchers. However, it lacks the rationale of Gemini Ultra, it is far more powerful.

At Google I/O 2024, users expect to see some improvement in this issue with 1.5 version upgrades to each of the free models in the family. They will most likely be made available to the Gemini chatbot and the Android Assistant.

Read more: Google I/O 2024 latest updates — Watch livestream of mega tech conference

New AI features at Google I/O

Google has previously previewed a new version of Gemini, which uses the voice assistant and video capabilities to describe and help with what's happening from the user’s camera perspective.

Gemini is capable of far more than what chat or voice interfaces presently allow. This includes consuming video and audio information. Gadinsider believes that both will be updated to include these additional input possibilities during I/O.

In addition, one of the most advantageous elements of Gemini over ChatGPT is its extensive interaction with the Google ecosystem. This includes search, flight maps, all of the user's papers, and YouTube, which are all accessible through extensions. YouTube Music has joined the list of extensions, but only in Gemini's Android assistant version.

Whereas, it's improbable, that Google may add third-party suppliers to the extension list. This would replicate capabilities seen in ChatGPT and Microsoft Copilot. If Google integrates this, firms like Uber and Kayak may have access to Gemini. If this occurs, people can plan a vacation and handle all bookings from within chat in Assistant.

Read more: Google Pixel 8a launched — Check out the latest specs and pricing

Google vs the AI competition

In terms of artificial intelligence, the world is transitioning from text to speech. This is evident in the fact that every AI lab is focusing on synthetic speech solutions.

Gadinsider also transitions from conversation to agents, in which users order the AI to complete a series of activities on their behalf rather than just having a casual discussion.

Moreover, this is something users have already seen from OpenAI. Apple is also rumoured to be considering this as a strategy for Siri 2.0, which users expect to debut at WWDC 2024 next month. To some extent, Google is trying similar things with the Gemini Assistant for Android.