Written by 6:35 am Generative AI, Latest news

### Enhanced Learning Capabilities Unveiled in Gemini 1.5 Pro

The latest for Google’s large language models.

Gemini 1.5 Pro Unveiled in Latest Update

Authored by Emilia David, an AI specialist writer. Before her tenure at The Verge, she delved into the intersection of technology, finance, and commerce.

Illustration of Google’s wordmark, written in red and pink on a dark blue background.

The latest iteration, Gemini 1.5 Pro, by Google introduces a revolutionary feature. It now has the capability to transcribe audio files without the need for manual input, extracting valuable insights from sources such as earnings calls or music embedded in videos.

During the Google Next event, Google announced the public release of Gemini 1.5 Pro on their AI software development platform, Vertex AI. The initial announcement of Gemini 1.5 Pro was made back in February.

Positioned as the intermediary model within the Gemini lineup, this new Gemini Pro variant outperforms its predecessor, the Gemini Ultra, in terms of functionality. Google asserts that Gemini 1.5 Pro can comprehend intricate instructions without the necessity for model adjustments.

Access to Gemini 1.5 Pro is exclusively available through Vertex AI. Currently, users primarily engage with Gemini chatbots, with the advanced version powered by Gemini Ultra. However, the sophistication of Gemini 1.5 Pro surpasses that of Gemini Ultra, enabling it to comprehend elaborate commands effectively.

Notably, Google has upgraded Imagen 2, the text-to-image generation model that enhances Gemini’s image creation capabilities, to include inpainting and outpainting functionalities. These features empower users to add or remove elements from images. Additionally, Google has integrated the SynthID online watermarking feature into all images generated using Imagen designs, enabling easy tracking through visible watermarks.

Some of the latest features of Imagen, particularly inpainting and outpainting, have been adopted by other text-to-image models like Getty’s Generative AI and Stability AI’s Steady Cascade by iStock. Moreover, these features are gaining traction among users of recent Samsung Galaxy devices.

Google is also exploring a feature that allows its AI to provide real-time information when queried through Google Search. While AI-generated responses can sometimes lack accuracy, Google has taken precautions to prevent Gemini from responding to queries related to the 2024 US election.

Gemini recently faced criticism for generating misleading images of individuals.

Visited 2 times, 1 visit(s) today
Tags: , Last modified: April 11, 2024
Close Search Window
Close