Every week, Quartz rounds up news on product launches, updates, and funding from startups and companies focused on artificial intelligence.
Here’s what’s happening this week in the ever-evolving AI industry.
Nvidia’s AI audio model “Fugatto”
Nvidia Headquarters – Image: Nvidia
Nvidia (NVDA) this week announced a new AI audio model, Fugatto. It can use any combination of text and audio files to produce or convert “any combination of music, speech, and sounds described in a prompt.”
According to Nvidia, Fugatto stands for Foundational Generative Audio Transformer Opus 1.
With the new model, users can enter text prompts to generate music snippets, remove or add instruments from existing songs, change vocal accents and emotions, and even create sounds like you’ve never heard before. You can “generate” or “generate”.
“Fugatto is the first fundamental generative AI model that exhibits the ability to combine emergent properties (capabilities that result from the interaction of different trained abilities) with free-form instructions,” Nvidia said.
Claude’s new features
Claude’s Style Features – Image: Anthropic
Anthropic announced new updates to its Claude AI chatbot this week.
The new Google (GOOGL) Docs integration allows Claude to use the context of a user’s documents to improve the relevance and accuracy of responses, Anthropic said. Claude can summarize long documents and “see historical context” from Google Docs files.
This integration is available to Claude Pro, Team, and Enterprise users.
Claude’s new styles feature allows users to customize how the chatbot responds according to their preferred communication style and work requirements. Users can choose formal, concise, or descriptive styles. Claude can also generate custom styles from uploaded sample content. Anthropic also added global profile settings to its chatbots, giving them the ability to tell them “things you want them to keep in mind,” such as their preferred coding language.
AI agent startup /dev/agents raises $56 million
/dev/agents website – Screenshot: /dev/agents
AI agent startup /dev/agents came out of stealth mode this week, announcing a $56 million seed round. The funding round was co-led by Index Ventures and CapitalG.
The startup, which is building an operating system for AI agents, is led by former Stripe chief technology officer David Singleton and former Google and Meta (META) vice presidents Hugo Barra, Ficus Kirkpatrick and Nicholas Jitkoff. was founded by.
“Today you can build an AI demo in a matter of hours, but it’s nearly impossible for consumers to actually get something they can trust with their credit card,” Singleton says. “Just as Android made mobile development accessible to nearly every developer, we are building a platform that will help bring AI agents mainstream.”
Runway image generation model “Frames”
Runway CEO Cristóbal Valenzuela and artist and musician Claire L. Evans attend the 2024 AI Film Festival Los Angeles on May 1, 2024 in Los Angeles, California – Photo by Tommaso Boddi (Getty Images)
Runway this week announced Frames, a new base model for image generation. The AI startup, which develops multimodal AI systems for video, image, and audio generation, said the new model represents “a major step forward in style control and visual fidelity.”
According to Runway, frames can maintain stylistic consistency when generating images. Access to Frames is deployed through the Gen-3 Alpha foundation model and Runway’s API.
Luma’s first image generation model
Luma AI Dream Machine – Image: Luma
Visual AI platform Luma this week released its first image generation model, the Luma AI Photon Image Model. The underlying text-to-image model was built on Luma’s Universal Transformer architecture. Luma says this model is 800% faster and cheaper than other models.
Luma also announced a free subscription-based offering of its Dream Machine visual AI service on the web and iOS. The Dream Machine model was released in June and has 25 million registered users.
For the latest news, follow us on Facebook, Twitter and Instagram.
