Create articles from any YouTube video or use our API to get YouTube transcriptions
Start for freeAnthropic's Claude Rolls Out Exciting New Features
Anthropics Claude has introduced several cool new features that enhance its functionality and user experience:
Projects Feature
Claude now allows users to create "projects", which function similarly to custom GPTs on ChatGPT. With projects, you can:
- Upload multiple files for context
- Use that uploaded information consistently across chat sessions
- Provide custom instructions for handling prompts
- Share projects with team members
Updated User Interface
The Claude interface has received some updates:
- Chat box and model selection remain in the main view
- Projects can be selected from a dropdown menu
- Recent chats are visible
- Profile button moved to bottom left
- Sidebar pops out when hovering on left side, showing recent and starred chats
Artifacts Feature
A new right-side window displays code output and other artifacts generated during conversations with Claude.
How to Use Projects
To create and use a project in Claude:
- Click "Create Project" in top right
- Name the project and add a description
- Set custom instructions if desired
- Upload relevant files (up to 30MB total)
- Ask questions about the uploaded content
Projects allow Claude to maintain context across multiple chat sessions, making it more efficient for ongoing research or analysis tasks.
Google's Gemini 1.5 Pro Boasts Massive Context Window
Google has announced some major updates to their Gemini AI model:
2 Million Token Context Window
Gemini 1.5 Pro now has an enormous 2 million token context window. To put this in perspective:
- It can handle about 1.5 million words of combined input and output
- This surpasses the entire Harry Potter series (1.1 million words)
- It nearly matches the entire A Song of Ice and Fire series (1.7 million words)
While most users may not need such a large context window, it represents a significant technological advancement.
New Capabilities
- Code execution: Gemini 1.5 Pro can now execute code
- 110 new languages added to Google Translate using AI
Gemma Open Source Model
Google released Gemma, a new open source model:
- Available in 9B and 27B parameter versions
- Outperforms Llama 3 and Grok on most benchmarks
Gmail Integration
Google is rolling out a Gemini sidebar in Gmail for select users:
- Summarizes conversations
- Shows unread emails
- Helps draft emails
- Offers additional suggestions
This feature is currently available to Gemini Business, Enterprise, Education, or Google One AI Premium subscribers.
OpenAI's Latest Developments
OpenAI has been busy with several new announcements and partnerships:
Advanced Voice Assistant
The highly anticipated voice assistant demoed a few weeks ago is nearing release:
- Originally planned for late June
- Delayed by one month to improve content moderation and user experience
- Some users report early access, including background sound effects in narrations
Critic GPT
OpenAI introduced Critic GPT, a new model based on GPT-4 designed to catch errors in ChatGPT's code output:
- Reviews code generated by ChatGPT
- Identifies potential issues and security vulnerabilities
- Aims to improve overall code quality
Mac Desktop App
The ChatGPT Mac desktop app is now available for all users, including free tier:
- Quick access using Option + Spacebar shortcut
- Functions similarly to Spotlight search
Acquisition of Multi
OpenAI purchased Multi, a remote collaboration platform:
- Allows screen sharing and collaborative work
- Potential integration with ChatGPT for interactive problem-solving
Partnership with TIME Magazine
OpenAI and TIME announced a multi-year content partnership:
- OpenAI gains access to TIME's 101-year archive
- TIME will use OpenAI technology to develop new products
This partnership represents a shift in media company attitudes towards AI, with many now seeking collaborations rather than opposing the technology.
Other Notable AI News
Toys R Us AI Commercial
Toys R Us created a new commercial using OpenAI's Sora:
- Mixed reactions from viewers
- Debate over AI's role in advertising and content creation
11 Labs Reader App and TIME Partnership
11 Labs introduced a new reader app and partnered with TIME magazine:
- Reader app converts any text to high-quality audio on iOS (Android coming soon)
- TIME integrating 11 Labs technology for audio versions of articles
Character AI Voice Calls
Character AI now allows voice calls with AI avatars:
- Popular among younger users
- Enables phone-like conversations with fictional characters
AI Music Copyright Issues
Record labels are suing AI music tools like Suno and Yudo:
- Claim unauthorized use of copyrighted music for training
- Raises questions about fair use and the future of the music industry
YouTube's AI Music Licensing Efforts
YouTube is negotiating with major record labels:
- Offering lump sum payments for AI training rights
- Aims to avoid legal issues faced by other AI music tools
Figma's AI-Powered Design Tools
Figma announced new AI features at their Config conference:
- Text-to-design capabilities
- Auto-generated placeholder text
- Enhanced asset search
- Currently in beta, will become a paid feature
Synthesia's AI Avatar Improvements
Synthesia updated their AI avatar technology:
- More expressive avatars showing various emotions
- Personal avatars that look and sound like the user
- Full-body avatar options
New AI Video Editing Tool from Captions
Captions launched an AI video editing tool:
- Adds backgrounds, text, and animations to existing videos
- Simulates green screen effects
Magnific's Relight Feature
Magnific, an AI image upscaler, introduced a new relight feature:
- Reimagines images in different settings or styles
- Uses text prompts or reference images
Stability AI's New CEO
Stability AI appointed a new CEO:
- Former CEO of Weta Digital
- Brings experience from major film and TV productions
Rabbit R1 Security Concerns
A security audit revealed significant issues with the Rabbit R1 device:
- Exposed API keys found in the code
- Potential for data breaches and unauthorized access
Etched's Specialized AI Chip
Etched announced a new chip designed for transformer models:
- Claims to be 20 times faster than Nvidia H100s
- Potential competitor to Grok's chips
Meta's AI Chatbots on Instagram
Meta began testing user-created AI chatbots on Instagram:
- Allows individuals to create custom chatbots
- Will be clearly labeled as AI
AI-Powered Olympics Coverage
The upcoming Olympics will feature AI-generated content:
- Personalized feeds based on user interests
- AI version of commentator Al Michaels providing recaps
As AI continues to evolve at a rapid pace, these developments showcase the technology's growing impact across various industries and applications. From improved language models and creative tools to new hardware and content delivery methods, AI is reshaping how we interact with technology and consume information.
Stay tuned for more updates as the AI landscape continues to evolve and new innovations emerge.
Article created from: https://youtu.be/ia7exPLf_do?feature=shared
