AI News

Stable Audio 2.0 by Stability AI enhances music generation, allowing users to create up to three-minute tracks from text prompts. Its user-friendly design and versatile input options enable unique compositions, yet the music often lacks emotional depth and features odd vocalizations, limiting its authenticity.

OnePlus announces its AI Eraser tool for photo editing, enabling users to remove unwanted objects easily. Available on several models, this feature uses a proprietary AI model for seamless background generation. More AI features are expected in the future.

Microsoft's Bing and Google's SGE raise concerns for businesses about web traffic decline due to AI. However, Perplexity's traffic has surged 40% month-over-month. BrightEdge advises continued investment in SEO, focusing on broader topics and structured content. AI search's evolution demands adaptable strategies for marketers.

Assembly AI has launched Universal-1, a speech recognition model trained on 12.5 million hours of multilingual audio. It boasts improved accuracy, reduced hallucinations, and faster processing times—transcribing one hour of audio in 21 seconds. Universal-1 is designed for diverse applications, now available via API.

Meta's image generator struggles with accurately depicting mixed-race couples, often producing stereotypical images. Despite numerous attempts, prompts for an East Asian man and Caucasian woman mostly resulted in images of two Asians. This reinforces societal biases in AI.

OpenAI has enhanced ChatGPT's image generation by adding an interactive paintbrush tool that allows users to modify images, powered by the DALL-E 3 model. Users can adjust sections of images and apply preset styles, amidst ongoing legal challenges and criticism regarding copyright issues in AI art.

Resemble AI has launched Rapid Voice Cloning, enabling users to create voice clones from just a 10-second to one-minute audio sample in about a minute. This feature enhances accessibility and personalization in various applications, including content creation and virtual assistants. Users must subscribe for additional clones, starting at $29 monthly.

Cognition's AI software engineer, Devin, can autonomously write code, benefiting developers but raising security concerns in sensitive industries. Codium AI's "Codiumate" offers semi-autonomous support, enhancing workflows while ensuring compliance. Powered by advanced models, it maintains high security standards and has gained 500,000 users.

The panel at Nvidia GTC 24 discussed digital twins and generative AI's role in transforming industries, emphasizing the need for standardized assets like OpenUSD. Key experts shared insights on digitalization, simulation accuracy, and collaborative technologies driving efficiency in manufacturing.

Gravity Jack, a Spokane-based machine learning company, launched Investobot, an AI-powered chatbot to facilitate real-time investor interactions. As they seek $3.5 million for the mobile game WarTribe of Binyamin, Investobot streamlines due diligence, enhancing engagement and trust in the investment process.

AI in cybersecurity offers both significant opportunities and challenges. While many security professionals see AI as a tool to augment their roles, there’s a disconnect in understanding between executives and staff. AI aids in threat detection and automates tasks, but its misuse by attackers raises concerns.

Stability AI has launched Stable Audio 2.0, enhancing its generative AI offerings. The updated model allows users to create high-quality audio tracks up to three minutes long, supports audio-to-audio generation, and emphasizes copyright compliance. The company aims to innovate while maintaining IP rights.

Amazon has unveiled its new Titan Image Generator, positioning itself alongside numerous tech giants and startups that have already ventured into the realm of image creation.

Walmart is embarking on an exciting journey with a series of innovative technology experiments aimed at transforming the shopping experience for its customers. The focus is on integrating generative AI to enhance convenience and personalization through various features, such as an interior design assistant, a smart search tool, and a voice-command shopping assistant.

Salesforce has launched the public beta of Einstein Copilot, a conversational AI assistant aimed at enhancing user interaction with data and workflows. Built on the Hyperforce architecture, it integrates generative AI capabilities while emphasizing data context. CEO Clara Shih highlights its unique metadata structure and ongoing relevance amid the AI trend.