🗞 ChatGPT 4o-mini Quietly Gains Multimodal Features 🚀

Key AI Developments from the Last 24 Hours

TGIF, enthusiasts! 🌟 Digitize Dispatch brings you the latest, most impactful AI news, cutting through the noise. No filler, just the updates driving the future of AI.

🔎 The Latest on the AI Frontier:

  • ChatGPT 4o-mini Quietly Gains Multimodal Features 🚀

  • OpenAI CEO Addresses Restructuring Rumors 🗣️

  • Musk's Companies Set for Massive AI Investment in 2024 💻

  • LinkedIn Scales Back AI Features After User Feedback 🔙

  • Nvidia's Next-Gen GPU Rumored to Have Impressive Specs 🎮

  • Meta's Advanced Llama 3.2 Models Now on AWS 🧠

  • Anthropic's Revenue Projected to Skyrocket in 2024 📈

  • AI Model Identifies Drug Candidates for Thousands of Diseases 💊

  • AI-Generated Explicit Content Becoming a School Issue 🚨

  • Other news you might find interesting 👀

🚀 ChatGPT 4o-mini model quietly gains multimodal capabilities, enhancing user experience. Link

  • The upgraded model now supports image generation, code writing, image description, and memory interaction, matching the full 4o model's functionality.

  • This significant improvement is currently available only to Plus subscribers, offering faster access to various ChatGPT tools.

  • The silent rollout represents a major leap from the previous 3.5 model, potentially revolutionizing user interactions with AI assistants.

💼 OpenAI's Altman denies plans for "giant equity stake" amid company restructuring talks. Link

  • At an all-hands meeting, CEO Sam Altman refuted rumors of receiving a large equity stake in OpenAI, stating there are "no current plans."

  • OpenAI's board is considering restructuring the company to a for-profit business, while maintaining a separate nonprofit entity.

  • Recent executive departures, including CTO Mira Murati, are reportedly unrelated to the potential restructuring, according to Altman.

💰 Tesla and xAI projected to spend $20B on AI in 2024, per Morgan Stanley. Link

  • Tesla is expected to invest $10B in AI projects, including a supercluster with 50,000 NVIDIA H100 GPUs and 20,000 Tesla D1 chips.

  • Elon Musk's xAI is also projected to spend $10B, primarily on its Colossus supercluster with 100,000 H100 GPUs for training the Grok 3 LLM.

  • This massive AI investment by Musk's companies highlights his strategy to secure a leading position in the rapidly evolving AI landscape.

🤖 LinkedIn scales back AI initiatives amid user feedback and technical challenges. Link

  • LinkedIn is rolling back some of its artificial intelligence features after user complaints and technical issues.

  • The company had introduced AI-powered tools for job listings, profile summaries, and messaging, but is now reassessing their implementation.

  • This move highlights the challenges tech companies face in balancing AI innovation with user experience and preferences in professional networking platforms.

🎮 Nvidia's RTX 5090 rumored to have 32GB VRAM and 600W power draw. Link

  • The upcoming GeForce RTX 5090 is reportedly planned to include 32GB of GDDR7 memory and 21,760 CUDA cores.

  • Power requirements are rumored to increase significantly, with the RTX 5090 potentially drawing 600 watts, up from RTX 4090's 450 watts.

  • The RTX 5080 is also said to see a power increase to 400 watts, with 16GB GDDR7 memory and 10,752 CUDA cores.

🧠 Meta's advanced Llama 3.2 models now available on AWS, expanding AI capabilities. Link

  • New offerings include multimodal vision models (11B and 90B), lightweight edge-optimized models (1B and 3B), and a content safety model (Llama Guard 3 11B Vision).

  • All models support 128K context length and 8 languages, with enhanced linguistic understanding and contextual data leveraging.

  • AWS customers can easily access and deploy these models using services like Amazon Bedrock, SageMaker, and EC2, with built-in responsible AI features.

💰 Anthropic's revenue projected to soar to $1 billion in 2024, marking 1,000% YoY growth. Link

  • 60-75% of Anthropic's revenue is expected to come from third-party API access, highlighting strong demand for its AI capabilities.

  • The company is reportedly in talks for a new funding round that could value it between $30-40 billion, potentially doubling its current valuation.

  • Anthropic recently launched an enterprise plan featuring a 500K context window and enhanced security, positioning itself as a key player in the rapidly evolving AI market.

🧬 AI model TxGNN identifies drug candidates for 17,000+ diseases, including rare ones. Link

  • Harvard Medical School researchers developed TxGNN, an AI tool that can repurpose existing drugs for over 17,000 diseases, many without current treatments.

  • The model is 50% better at identifying drug candidates and 35% more accurate in predicting contraindications compared to leading AI models for drug repurposing.

  • TxGNN is freely available and aims to accelerate therapy discovery for rare and neglected conditions, potentially reducing health disparities.

🚨 AI-generated child abuse imagery surges in schools, affecting millions. Link

  • 15% of high schoolers reported hearing about sexually explicit AI-generated "deepfakes" of someone from their school in the past year.

  • The Center for Democracy and Technology warns that generative AI has increased opportunities for both victimization and perpetration among students.

  • Experts are only now beginning to grapple with the dark side of AI's image generation capabilities and its impact on children's safety and privacy.

More news you might find interesting:

  • 🤝 UK regulator approves Amazon's $4B AI partnership with Anthropic, citing lack of jurisdiction. Link

  • 📱 Apple users irked by lack of camera improvement across 4 iPhone generations. Link

  • 💧 AI's water use pales in comparison to agriculture, especially livestock feed. Link

  • 🎁 Samsung unveils Galaxy S24 FE with AI features at $650, bridging budget and premium offerings. Link

  • 💼 ByteDance secures record-breaking $10.8 billion corporate loan in Asia. Link

  • 🖼️ Grok on X enhances user experience with new image gallery and vision capabilities. Link

  • 🎨 Napkins.dev leverages Llama 3.2 AI to transform wireframes into code instantly. Link

  • 🚔 DoJ warns companies they'll be held accountable for AI-related crimes. Link

  • 🧠 Larger AI language models more likely to give wrong answers than admit ignorance, study finds. Link

  • 🎬 Runway allocates $5M to fund 100 films using AI-generated video. Link

What did you think of today's newsletter?

Login or Subscribe to participate in polls.

Have any feedback? Send us an email