DMR News

Advancing Digital Conversations

From Bard to Gemini: Google’s AI Chatbot Gets a Powerful Upgrade

ByHuey Yee Ong

Feb 16, 2024
From Bard to Gemini: Google's AI Chatbot Gets a Powerful Upgrade

Google recently updated its AI chatbot, Bard, renaming it to Gemini, marking a significant milestone. Launched nearly a year ago, Bard initially faced criticism for various shortcomings. However, through continuous enhancements, including two major updates to its underlying Large Language Model (LLM) and several incremental improvements, Bard has transcended its initial limitations.

Rebranding Bard to Gemini

On February 8th, 2024, Google announced an ambitious rebranding, shifting from Bard to Gemini, to underscore the advanced technology powering the chatbot. Gemini represents Google’s latest and most sophisticated LLM, engineered to cater to a wide array of tasks with its three distinct versions:

  • Gemini Nano: Designed for lightweight applications, offering efficiency and speed for basic tasks.
  • Gemini Pro: The version currently powering Google Bard, optimized for balanced performance across a broad range of more complex queries.
  • Gemini Ultra: The most advanced version, capable of handling the most demanding tasks with superior processing power and enhanced capabilities.

The rebranding aligns the chatbot’s identity with its core technology, perhaps a nod to Microsoft’s strategy with Bing Chat’s transformation into Copilot, which similarly introduced a spectrum of service levels. This rebranding is more than a mere name change; it symbolizes Google’s commitment to aligning product names with the technology that underpins them, albeit at the risk of some confusion due to the dual use of the Gemini brand.

Gemini’s Reach: Beyond Android to iOS

The introduction of the Gemini app for Android devices is a pivotal component of this rebranding strategy. Slated for release next week, the app promises an accessible mobile interface for engaging with Gemini, integrating the convenience of on-the-go queries with the functionality of Google Assistant. This development aims to streamline access to Gemini’s capabilities, allowing users to interact through voice commands or text inputs across various activation points, including power button presses and corner swipes. The Gemini app is not limited to Android users; iOS users can anticipate accessing Gemini through the Google app in the near future. This inclusivity ensures that a broader audience can experience the advanced AI chatbot, further extending Google’s reach in the AI space.

Introducing Google One AI Premium Plan

In tandem with these advancements, Google introduced a new premium subscription model, the Google One AI Premium Plan, which includes:

  • Access to Gemini Advanced: Subscribers can use the more capable version of Gemini, similar to ChatGPT Plus, for advanced inquiries and tasks.
  • 2TB of Storage: A significant amount of cloud storage to accommodate users’ extensive data needs.
  • Latest Google Innovations: Subscribers gain early or exclusive access to new features and advancements from Google, ensuring they are always at the cutting edge of technology.

At $19.99 per month, this strategic move positions Google to capitalize on the growing demand for enhanced AI services, offering significant value through advanced features and increased storage capacity.

Unveiling Gemini 1.5: The Next-Generation AI Model

The unveiling of Gemini 1.5, a next-generation AI model, underscores Google’s relentless pace of innovation. Despite Gemini 1.0’s recent debut in December, the 1.5 iteration boasts substantial improvements, including:

  • Expanded Context Window: Allows the model to process and understand larger blocks of text at once, significantly improving comprehension and the ability to handle complex queries.
  • Refined Performance Metrics: Enhances the model’s efficiency, ensuring faster and more accurate responses to user prompts.
  • Updated Mixture-of-Experts (MoE) Framework: This architectural innovation enables Gemini 1.5 Pro to selectively activate the most relevant pathways in its neural network, improving learning efficiency and performance.
  • Increased Token Capacity: In a limited preview available to developers and enterprise customers, Google offers the capability to process up to one million tokens for free through AI Studio and Vertex AI, a substantial leap from the 32,000 token capacity of its predecessor.

Google CEO Sundar Pichai highlighted the model’s efficiency and the promise of new capabilities facilitated by longer context windows, indicating a future where AI can tackle increasingly complex tasks with greater nuance.

Enhanced Performance and Efficiency

Gemini 1.5’s architecture, based on an updated Mixture-of-Experts (MoE) framework, enhances its learning efficiency by activating the most relevant neural network pathways. This architectural innovation allows Gemini 1.5 Pro to process up to one million tokens, a substantial leap from the 32,000 token capacity of its predecessor. This increase in token capacity significantly enhances the model’s ability to absorb and process information, setting a new standard for AI performance.

Safe and Responsible Deployment

Google demonstrated Gemini 1.5 Pro’s capabilities through various tests, including processing a silent film and responding to multimodal queries with remarkable accuracy. The model’s performance in standard benchmarks and specialized evaluations like the Needle In A Haystack (NIAH) and Machine Translation from One Book (MTOB) further attest to its advanced learning capabilities and potential for diverse applications.

To address potential concerns about the deployment of such powerful technology, Google has undertaken comprehensive evaluations to ensure the safe and responsible introduction of Gemini 1.5 Pro. This cautious approach reflects a broader industry imperative to balance innovation with ethical considerations and user safety.

Moving Forward

As Google prepares to offer Gemini 1.5 Pro to developers and enterprise customers in a limited preview, the AI community awaits the wider implications of this technological leap. With plans to introduce variable pricing tiers based on token context windows, Google is not only advancing the frontiers of AI technology but also shaping the commercial landscape of AI services.

In conclusion, Google’s rebranding of Bard to Gemini and the subsequent introduction of Gemini 1.5 represent a bold stride forward in AI development. By marrying innovative technology with strategic market positioning, Google continues to redefine the boundaries of what AI can achieve, promising an era of more intuitive, accessible, and powerful AI-driven solutions.

Featured Image courtesy of GOOGLE

Huey Yee Ong

Hello, from one tech geek to another. Not your beloved TechCrunch writer, but a writer with an avid interest in the fast-paced tech scenes and all the latest tech mojo. I bring with me a unique take towards tech with a honed applied psychology perspective to make tech news digestible. In other words, I deliver tech news that is easy to read.