AI This Week: Secure Emails, Olympic Games, and Creative Enhancements

This week’s AI roundup features ProtonMail’s commitment to secure communications with Proton Scribe, Meta’s ambitious expansion with its Llama AI model, and Adobe’s creative revolution through Firefly AI tools. Additionally, Google’s Gemini AI is set to transform the Paris Olympics experience, while Vena Insights redefines business analytics with sophisticated AI-driven tools.

ProtonMail’s Privacy-First AI Assistant

Proton, known for its suite of privacy-centric apps, has launched Proton Scribe, a new AI-enabled writing assistant designed to operate directly on users’ devices, enhancing email privacy and security. Unlike traditional AI tools that store data on cloud servers, Proton Scribe ensures all data remains on the local device, aligning with the needs of privacy-conscious enterprises.

Proton Scribe is built on the open-source Mistral 7B language model and is available under the GPL-3.0 license, facilitating third-party security audits. While the tool focuses on privacy, it does not use local data to personalize responses, which might result in more generic outputs. However, Proton has integrated editing features to improve usability.

Featured Image: Proton Mail

Currently available on Proton Mail’s web and desktop platforms for $2.99 monthly for business users, Proton plans to extend its availability to mobile devices and other products depending on user demand.

Meta Unveils Its Largest Open AI Model Yet

Meta has launched its most expansive open-source AI model to date, Llama 3.1 405B. With 405 billion parameters, this model is available for use on cloud platforms such as AWS, Azure, and Google Cloud and is currently powering chatbots on WhatsApp and Meta.ai for U.S. users.

The Llama 3.1 405B, trained on 16,000 Nvidia H100 GPUs, incorporates newer development techniques, positioning it on par with top proprietary models like OpenAI’s GPT-4o and Anthropic’s Claude 3.5 Sonnet. Meta has enhanced the model’s dataset to include a richer mix of non-English and recent web data to bolster its capabilities, improving performance across diverse tasks. However, despite these advancements, the model’s use of training data remains undisclosed due to competitive and legal reasons.

Meta has also introduced smaller models, Llama 3.1 8B and Llama 3.1 70B, suitable for general-purpose applications like powering chatbots and code generation. Like the larger Llama 3.1 405B, these models can integrate third-party tools and APIs to extend their functionality.

Featured Image: Meta

Adobe Enhances Creative Suite with Firefly AI Tools

Adobe has introduced new Firefly AI tools for Photoshop and Illustrator, allowing users to generate detailed textures and images through simple text prompts, significantly reducing the time required for complex design tasks.

In Illustrator, the new “Generative Shape Fill” tool, currently in beta, lets users add textures and details to shapes using either text prompts or style references. This feature is powered by the beta version of Adobe’s Firefly Vector model, allowing for precise control over the level of detail added. Unlike many AI generators, Illustrator provides flexibility for manual adjustments post-generation.

Featured Image: Adobe

Photoshop has broadened access to its Firefly text-to-image generator, enabling users to create detailed images directly within the app. This feature, driven by the Adobe Firefly Image 3 Foundation model, simplifies transforming textual descriptions into rich, detailed visuals.

Adobe is also committed to a creator-friendly approach by incorporating these AI tools into the existing Creative Cloud subscriptions without additional costs, offering a generous allotment of generative credits monthly. This strategy retains existing customers and encourages upgrades to more premium plans, as evidenced by a reported increase in conversions since Firefly’s initial launch in March 2023.

Furthermore, Adobe continues to innovate beyond AI with new tools in Illustrator, such as the Dimension Tool for measuring image metrics and an improved Mockup tool for applying logos realistically to any product image. Photoshop also sees updates with the new Selection Brush Tool and Adjustment Brush Tool to streamline editing tasks.

Google’s Gemini AI Dominates the Paris 2024 Olympics Broadcast

Google is set to dramatically influence the upcoming Paris 2024 Olympics as the “official AI sponsor for Team USA.” In a notable collaboration with NBCUniversal, Google will showcase its suite of AI tools, prominently featuring Gemini AI throughout the broadcast.

During the games, viewers can expect to see  Google Maps’ 3D Immersive Views, offering photorealistic representations of key venues like Versailles and the Stade Roland Garros. This integration aims to enhance the viewer experience by providing a detailed visual understanding of where each event unfolds.

Moreover, the broadcast will cleverly incorporate Google Search AI Overviews. Commentators will use this feature to instantly provide in-depth answers to complex Olympic and Paralympic queries, showcasing the AI’s ability to handle live, nuanced information.

Featured Image: Google

Additionally, selected Olympic and Paralympic athletes will feature in social media videos and late-night promos, utilizing tools like Google Lens and Circle to Search to interact with the city of Paris, further demonstrating the practical uses of Google’s AI technologies.

Tool in Spotlight: Vena Insights

In this week’s spotlight, explore Vena Insights, a dynamic intelligence reporting and analytics solution embedded with Power BI and Microsoft’s state-of-the-art AI and machine learning technologies. Designed to streamline business planning processes, Vena Insights integrates directly with Vena data, offering real-time access for efficient and intelligent reporting.

Transform Data into Actionable Intelligence

Vena Insights enhances decision-making across your organization by enabling quick, efficient data transformations into actionable insights. The platform supports self-service analytics, which lightens the load on finance teams and empowers employees to generate and share reports easily. The tool’s sophisticated AI capabilities allow users to customize machine learning models to uncover key business drivers, analyze risks, and track emerging trends swiftly.

Featured Image: Vena

Maximize Data Potential with Advanced Tools

Vena Insights extends beyond mere data analysis; it includes predictive analytics and anomaly detection to enhance forecast accuracy and pinpoint anomalies. Users benefit from the full spectrum of Microsoft’s AI advancements, allowing for a cutting-edge analytical experience. The platform also enables users to create personalized dashboards and utilize an extensive range of editable data visualizations for in-depth insights.

Keep ahead of the curve – join our community today!

Follow us for the latest discoveries, innovations, and discussions that shape the world of artificial intelligence.


Other Work

We should talk.

372 Richmond Street West, Suite 210
Toronto, ON M5V 1X6

+1.647.289.6838
info@trewknowledge.com