VENDORiQ: Google Bids to Democratise AI Adoption Through Gemini Offerings

Google unveils compelling functionalities for organisations looking to leverage the latest advancements in generative AI

The Latest

Google has announced significant advancements with its Gemini models. These include the unveiling of Gemini 1.5 Pro, its most capable generative AI model yet; launching Gemini Code Assist, an enterprise-focused AI code completion and assistance tool; and integrating Gemini into Threat Intelligence.

Why It’s Important

The expanded access to powerful AI models like Gemini 1.5 Pro, now available in public preview on Vertex AI, which can handle up to 1 million tokens of context, enables businesses to leverage advanced generative AI capabilities more easily. This allows for more efficient content creation, data analysis, and problem-solving across various functions.

Gemini Code Assist, an enterprise-focused AI code completion and assistance tool, competes directly with GitHub’s Copilot Enterprise and aims to provide developers with accurate code suggestions, deeper insights, and streamlined workflows.

The integration of Gemini into Threat Intelligence enables businesses to automatically identify and understand online threats more effectively. A novel assisted investigation functionality transforms natural language into new detections, condenses event data into summaries, provides actionable recommendations, and guides users through the platform using conversational chat interactions.

The announcements are an example of the shape of things to come from AI. AI will consolidate back into being a collection of software services under a larger ecosystem of capabilities that support automation. Over time, AI will become ‘invisible’ to most users since it will be built into almost all software that manages processes (which is really all software).

Who’s Impacted

  • Developers
  • Data analysts
  • Cyber security practitioners

What’s Next?

Gemini is built with responsible AI principles in mind, but businesses must carefully evaluate the model’s outputs and ensure they are unbiased and factually correct. Tracking sources and connecting AI responses to their data will be important.

While Gemini 1.5 Pro boasts impressive capabilities, such as analysing long videos and transcripts, it may still struggle with certain complex tasks. For example, determining whether a video is AI-generated can be challenging for the model. Also processing large amounts of data, such as long videos or code repositories, can lead to increased latency and computational requirements. Businesses should be mindful of these performance limitations when deploying Gemini 1.5 Pro in their workflows.

Gemini 1.5 Pro is currently only available to a select group of developers and enterprise customers, with limited public access. Businesses should consider the implications of relying on a model that is not yet widely accessible.

Trouble viewing this article?

Search