What new features does Google’s Gemini AI bring to Chrome’s address bar?
Cover Photo Major News from Google's Gemini AI, Luma AI's Dream Machine 1.6, DeepMind's GenRM, Salesforce, Anthropic, Canva, and Spotter

Chrome’s New Frontier: Gemini AI Joins the Address Bar

Google has integrated its Gemini AI chatbot directly into Chrome’s address bar, marking a significant shift in how users interact with the internet. By typing “@gemini” followed by a query, Chrome users can now access AI assistance without leaving their browser. This update, available in over 100 countries, utilizes Gemini 1.5 Flash, a lightweight version of Google’s advanced language model. While it lacks contextual awareness of browsing activity, it still offers powerful AI capabilities. This move aligns with Google’s strategy of incorporating AI across its products and could dramatically increase AI accessibility for millions of users worldwide. However, it also raises questions about data privacy and the growing influence of AI in digital life.

Luma AI Enhances Video Generation with Camera Motion Controls

Luma AI has introduced Dream Machine 1.6, an updated AI video generator that offers users greater control over camera movements. The new version includes 12 different camera motion options, such as panning, orbiting, and zooming, which can be easily accessed through a dropdown menu when typing prompts. This feature allows for more precise and realistic video creation, addressing a key limitation in AI-generated content. Early adopters have praised the update for its improved dynamic motion and time-saving capabilities. The enhancement puts Luma AI in direct competition with other AI video providers and demonstrates the ongoing advancements in AI-driven video production technology.

DeepMind’s GenRM: Enhancing LLM Accuracy Through Self-Verification

DeepMind has introduced GenRM, a novel approach to improve the accuracy of large language models (LLMs) by leveraging their generative capabilities for self-verification. Unlike traditional verifiers, GenRM trains models to generate and evaluate their own outputs, incorporating techniques like chain-of-thought reasoning. This method outperformed standard approaches in various reasoning tasks, including complex math problems. GenRM allows for flexible balancing of accuracy and computational costs, making it particularly useful for applications where current verification methods fall short. The approach unifies solution generation with verification, consistently improving performance across tasks and demonstrating favorable scaling with increased dataset size and model capacity.

Salesforce Acquires Tenyx to Bolster AI Voice Technology

Salesforce has announced its acquisition of Tenyx, a startup specializing in AI-powered voice agents. This strategic move aims to enhance Salesforce’s AI-driven solutions and marks a shift in the company’s recent stance on acquisitions. Tenyx, founded in 2022, serves clients across various industries, including e-commerce and healthcare. The deal, expected to close in the third quarter, will bring Tenyx’s co-founders and team into Salesforce. This acquisition aligns with similar moves by tech giants like Microsoft and Amazon, highlighting the growing competition for AI expertise. The purchase reflects Salesforce’s efforts to boost revenue growth through targeted acquisitions in the AI sector.

Anthropic to Release System Prompts for Artifacts, Enhancing AI Transparency

Anthropic has committed to releasing system prompts for its new Artifacts feature in the coming weeks, addressing concerns raised by researchers about incomplete information in their recent Claude family prompts disclosure. This move follows the company’s initial release of system prompts for its Claude models, which garnered praise for transparency. Artifacts, a recently launched feature allowing code snippet execution alongside Claude chats, was not included in the initial release due to its later public availability. The disclosure of system prompts is significant as it provides insight into model behavior and developer intentions, offering a rare glimpse into the configuration of large AI models.

Canva Raises Prices for Teams Plan, Citing AI Advancements

Canva, the popular design platform, has implemented significant price increases for its Teams product, particularly affecting customers on older pricing plans. The company attributes these changes to its expanding suite of generative AI tools, including Magic Studio. For some U.S. customers, the annual cost for a five-person plan has risen from $119.99 to $500, with a temporary discount offered. Australian users now face a per-user pricing model instead of a flat fee. Canva defends the decision, stating that the original pricing no longer reflects the product’s expanded capabilities. The move has sparked criticism from users, especially regarding the communication method. This price adjustment marks a shift from Canva’s initial positioning as an affordable design software alternative.

Spotter Unveils AI-Powered Creative Suite for YouTube Content Creators

Spotter has launched Spotter Studio, an AI-driven creative suite designed to assist YouTubers throughout their content creation process. The tool offers personalized suggestions for video concepts, thumbnails, and titles by analyzing a creator’s existing content and billions of top-performing YouTube videos. It also includes project management features for organizing tasks and collaborating with teams. While the suite aims to boost creativity and efficiency, its “Outliers” feature, which draws inspiration from other creators’ successful videos, has raised concerns about originality. Spotter claims early testing showed a significant increase in video views, but the tool’s impact on content diversity and authenticity remains to be seen.

Frequently asked questions

Google’s Gemini AI integration allows Chrome users to access AI assistance directly from the address bar by typing “@gemini” followed by their query. This feature is now available in over 100 countries and uses Gemini 1.5 Flash, a lightweight version of Google’s language model. While it can’t track browsing activity, it provides powerful AI capabilities without requiring users to leave their browser, making AI assistance more accessible to millions of Chrome users worldwide.
Luma AI’s Dream Machine 1.6 introduces 12 different camera motion controls, including panning, orbiting, and zooming options, accessible through a dropdown menu when creating prompts. This update significantly enhances the quality and realism of AI-generated videos by allowing users to precisely control camera movements. The improvement addresses previous limitations in AI video generation and makes the creation process more efficient and professional.
Canva’s new AI-powered features, including Magic Studio, offer enhanced design capabilities through generative AI tools. These features help users create more sophisticated designs, automate repetitive tasks, and access advanced creative options. However, these improvements have led to price increases, with the Teams plan now costing significantly more. The platform justifies these changes by pointing to the expanded functionality and value provided by their AI integration.
DeepMind’s GenRM improves AI accuracy by enabling large language models to perform self-verification of their outputs. The system trains models to generate and evaluate their own responses using chain-of-thought reasoning. This approach has shown superior performance in various reasoning tasks, particularly in complex mathematics, and allows for flexible balancing between accuracy and computational costs.
Canva has significantly increased prices for its Teams plan, with some U.S. customers seeing annual costs rise from $119.99 to $500 for a five-person plan. Australian users now face per-user pricing instead of flat fees. These changes reflect the platform’s investment in AI tools and expanded capabilities, though they’ve received criticism for their communication approach and the substantial price increase.
Spotter Studio is an AI-powered creative suite designed specifically for YouTube content creators. It analyzes billions of successful videos to provide personalized suggestions for video concepts, thumbnails, and titles. The tool includes project management features and an “Outliers” feature that draws inspiration from other creators’ successful content, though this has raised questions about content originality.
Dream Machine 1.6 distinguishes itself through its comprehensive camera motion controls, which many other AI video generators lack. The platform’s 12 different motion options provide more precise control over video creation, making it more competitive in the AI video generation market. Early users report improved dynamic motion and efficiency in video production compared to alternative solutions.
Picture of Gor Gasparyan

Gor Gasparyan

Optimizing digital experiences for growth-stage & enterprise brands through research-driven design, automation, and AI