🦾 Master AI & ChatGPT for FREE in just 3 hours 🤯

1 Million+ people have attended, and are RAVING about this AI Workshop.
Don’t believe us? Attend it for free and see it for yourself.

Save your spot here. (100 free spots only)

Highly Recommended: 🚀

Join this 3-hour Power-Packed Masterclass worth $399 for absolutely free and learn 20+ AI tools to become 10x better & faster at what you do

👉Save your seat  now (FREE for First 100)

🗓️ Tomorrow | ⏱️ 10 AM EST

In this Masterclass, you’ll learn how to:

🚀 Do quick excel analysis & make AI-powered PPTs 
🚀 Build your own personal AI assistant to save 10+ hours
🚀 Become an expert at prompting & learn 20+ AI tools
🚀 Research faster & make your life a lot simpler & more…

👉 Register here (Offer valid for First 100 people only)🎁

Today:

• Ideogram 2.0 Launches

• GPT-4o Fine-Tuning Now Available

• OpenAI and Condé Nast Collaborate

• Google AI Prompt Gallery

• D-ID Unveils AI Video Translation

• Microsoft Postpones Recall AI Launch

• Andreessen Horowitz Backs AI Copyright Startup

• Authors Sue Anthropic Over AI Training

AI News is getting INSANE! Mass Production Robots, OpenAI Rumor Mill and Unreal AI Video…

AI developments are accelerating, with Unitree Robotics announcing the G1 robot, which is agile, strong, and surprisingly affordable at $16,000. This advancement suggests a fully automated future might be closer than expected. 

In parallel, OpenAI's Sam Altman stirred rumors with a cryptic strawberry post, leading to speculation about upcoming AI breakthroughs. Despite the hype, no major new model has been released yet. 

Lastly, Runway ML showcased their Gen-3 Alpha Turbo, producing highly realistic AI-generated videos, underscoring the rapid progress in AI capabilities across various fields.

WATCH THE VIDEO ON YOUTUBE

Hands-on with Ideogram 2.0: The AI that makes text look incredible

Ideogram 2.0, a text-to-image AI model, is challenging established players like Midjourney and DALL-E 3 with its improved text rendering and customizable color palettes. The new features allow for creating precise and professional graphic designs, addressing a major issue in AI image generation. Ideogram's competitive pricing and the launch of a public beta API further expand its appeal to businesses and developers. 

While the technology democratizes design, it raises ethical concerns about the future of creative industries and the potential misuse of AI-generated imagery. Overall, Ideogram 2.0 signals a significant leap in AI-driven content creation.

VENTUREBEAT

Fine-tuning now available for GPT-4o

OpenAI has introduced fine-tuning for GPT-4o, allowing developers to customize the model for better performance and accuracy tailored to specific use cases. This feature, now available to all paid users, enables the model to follow complex instructions and adapt its tone. Early adopters, like Cosine and Distyl, have achieved state-of-the-art results in software engineering and SQL benchmarks using fine-tuned GPT-4o. 

The fine-tuning process is secure, with data privacy and safety measures in place, ensuring full control over business data. This launch is expected to significantly enhance the capabilities of AI applications across various domains.

OPENAI

OpenAI partners with Condé Nast

OpenAI has partnered with Condé Nast to integrate content from top brands like Vogue, The New Yorker, and Wired into its AI products, including ChatGPT and the new SearchGPT prototype. This collaboration aims to enhance news discovery by providing fast, reliable information with direct links to in-depth content. 

OpenAI is working with Condé Nast and other publishers to ensure the accuracy and integrity of AI-driven news delivery. Feedback from these partnerships will help refine and improve the integration of journalism into AI services, enhancing user experience in future updates.

OPENAI

Google debuts free ‘Prompt Gallery’ in AI Studio, supercharging developer tools

Google has introduced a new feature called Prompt Gallery in its AI Studio, enhancing tools for developers using the Gemini API. The gallery offers pre-built prompts for various applications, from coding to creative tasks. This free tool aims to democratize AI development by making advanced AI capabilities accessible to a broader audience. 

The update is seen as a strategic move by Google to capture a larger share of the AI tools market, offering significant value to developers and businesses while potentially reshaping AI development strategies.

VENTUREBEAT

D-ID launches an AI video translation tool that includes voice cloning and lip sync

D-ID has launched an AI video translation tool that translates videos into 30 languages while cloning the speaker's voice and synchronizing their lip movements. This tool, based on D-ID's previous success with animating photos, aims to help creators expand their reach globally by making video localization more accessible. 

The service is offered free to D-ID subscribers and competes with other AI-driven dubbing and video creation tools. With this technology, D-ID hopes to reduce localization costs and make professional video translation available to a broader audience, including smaller creators.

TECHCRUNCH

Microsoft’s Recall AI feature won’t be available for Windows testers until October

Microsoft's Recall AI feature, initially set to launch in June, has been delayed until October for Windows testers due to security concerns. Recall uses AI to capture and allow users to search through screenshots of nearly everything they do on their PC. The delay follows issues with database encryption and potential malware vulnerabilities. 

Microsoft is now making Recall an opt-in feature and enhancing its security measures, including database encryption and Windows Hello authentication. The full launch of Recall might be further delayed, depending on the outcome of these extended tests.

THE VERGE

Andreessen Horowitz leads $80 million bet on startup seeking to tame AI with copyright

A startup called Story, founded by S.Y. Lee and Jason Zhao, is addressing the challenge of AI potentially disrupting the economic landscape for creators by proposing a new intellectual property (IP) system. Story's platform uses blockchain technology to allow creators to quickly register their works, track usage, and collect royalties, aiming to protect creators' rights as AI systems increasingly utilize web content without permission. 

The startup has raised $80 million in a Series B funding round led by Andreessen Horowitz. Story’s approach is inspired by Creative Commons but adds monetization features to benefit creators in the AI age.

FORTUNE

Authors sue Anthropic for training AI using pirated books

A group of authors has filed a lawsuit against Anthropic, accusing the company of copyright infringement for using pirated books to train its AI models. The lawsuit, filed in California, claims that Anthropic used the Books3 dataset, part of "The Pile," which contains thousands of copyrighted ebooks from authors like Stephen King and Michael Pollan. 

The authors seek damages and a ban on Anthropic's use of copyrighted material. This lawsuit follows similar legal actions against other tech companies like Meta and Microsoft for allegedly using pirated content to train AI models.

THE VERGE

🧠RESEARCH

xGen-MM (BLIP-3): A Family of Open Large Multimodal Models

The xGen-MM (BLIP-3) framework introduces a series of open-source Large Multimodal Models (LMMs). It includes curated datasets, training methods, and model architectures. These models perform well in various tasks and focus on safety, reducing issues like hallucinations. The project aims to advance LMM research by providing open access to resources.

MeshFormer: High-Quality Mesh Generation with 3D-Guided Reconstruction Model

MeshFormer is a 3D reconstruction model that efficiently generates high-quality 3D meshes. Unlike previous methods, it uses 3D sparse voxels and combines transformers with 3D convolutions, guided by 3D structure and input normal maps. This approach reduces training complexity and improves mesh quality, making it effective for tasks like single-image-to-3D and text-to-3D conversions.

LongVILA: Scaling Long-Context Visual Language Models for Long Videos

LongVILA introduces a solution for training vision-language models on long videos. It features a new system, Multi-Modal Sequence Parallelism (MM-SP), which enables efficient long-context training on large-scale datasets. LongVILA improves video captioning and extends the number of frames models can handle, significantly boosting performance on long video tasks.

TableBench: A Comprehensive and Complex Benchmark for Table Question Answering

TableBench introduces a comprehensive benchmark for evaluating Large Language Models (LLMs) in table question answering (TableQA). Despite recent advancements, LLMs struggle with complex, real-world tabular data. The benchmark includes 18 fields across four categories, highlighting the gap between academic models and industrial needs. The study also presents TableLLM, which performs comparably to GPT-3.5, but experiments show significant room for improvement in meeting real-world demands.

TWLV-I: Analysis and Insights from Holistic Evaluation on Video Foundation Models

TWLV-I is a new video foundation model that excels in both appearance and motion understanding. It addresses the challenge of fairly evaluating video models by proposing a robust framework. TWLV-I shows significant improvements in action recognition benchmarks, outperforming existing models like V-JEPA, UMT, and DFN, demonstrating its superior video comprehension capabilities.

🛠️TOP TOOLS

Luma Dream Machine - AI model that makes high quality, realistic videos fast from text and images.

Frase - Empowers content creators to go from keyword to well-researched, SEO-optimized articles faster and better.

Trolly - Create professional SEO articles, 2x faster.

Flux-dev-lora-trainer - The easiest way to train ostris/flux-dev-lora-trainer is to use the form provided.

Evidently AI - Evaluate, test, and monitor your AI-powered products.

📲SOCIAL MEDIA

🗞️MORE NEWS

AI21 debuts Jamba 1.5, boosting hybrid SSM transformer model to enable agentic AI

AI21's new Jamba 1.5 model enhances AI capabilities by combining transformers with Structured State Space (SSM) models. The update includes JSON mode, function calling, and citation features, making it ideal for developing advanced AI systems. The models offer improved performance and transparency, supporting developers in building more sophisticated, agentic AI systems. VENTUREBEAT

AMD is going after Nvidia with a $5 billion acquisition

AMD is acquiring cloud computing firm ZT Systems for $4.9 billion to enhance its AI capabilities and compete with rivals like Nvidia. The acquisition aims to bolster AMD's AI infrastructure and accelerate large-scale AI deployment. The deal is expected to close by mid-2025 and contribute to profits by year's end. QUARTZ

Adobe drops ‘Magic Fixup’: An AI breakthrough in the world of photo editing

Adobe's new AI tool, Magic Fixup, revolutionizes photo editing by learning from video data, enabling sophisticated, automated adjustments while preserving artistic intent. This technology could transform industries from advertising to forensics. Adobe's decision to open-source the code marks a shift in its AI strategy, fostering collaborative innovation in digital creativity. VENTUREBEAT

Former Huawei ‘Genius Youth’ recruit launches humanoid robots to rival Tesla’s Optimus

Former Huawei "Genius Youth" recruit Peng Zhihui has launched Agibot, a robotics startup in Shanghai, unveiling a new generation of humanoid robots aimed at rivaling Tesla's Optimus. Agibot introduced five models, including the flagship Yuanzheng A2, and plans to ship 300 units by year-end. Peng aims to compete directly with Tesla in the growing humanoid robot market. SCMP

ElevenLabs’ text-to-speech app Reader is now available globally

ElevenLabs has globally launched its AI-powered text-to-speech app, Reader, which supports 32 languages and allows users to listen to text content like articles and e-books. The app includes hundreds of new voices and competes with similar apps like Speechify. ElevenLabs plans to add features like offline support and audio sharing soon. TECHCRUNCH

Google Cloud Run embraces Nvidia GPUs for serverless AI inference

Google Cloud has introduced Nvidia L4 GPUs to its Cloud Run serverless platform, enabling AI inference without the need for persistent cloud instances. This integration allows AI tasks to run only when needed, potentially reducing costs and improving efficiency. The platform supports popular AI models and aims to meet real-time processing demands while addressing performance concerns, particularly with cold start times. The pricing impact will vary based on usage patterns. VENTUREBEAT

Canva Steps Up To Challenge Adobe With AI Acquisition And Partnerships

Canva is stepping up its AI-powered design capabilities with the acquisition of Leonardo.Ai, a platform known for generating high-quality images, and a partnership with Getty Images, which integrates a vast stock photo library into Canva’s platform. These moves position Canva as a strong competitor to Adobe, especially in the social commerce and enterprise markets, by enhancing its AI tools and expanding its market reach. However, to truly challenge Adobe's dominance, Canva will need to develop a broader ecosystem that addresses various content creation needs beyond design. FORBES

SleekFlow snaps up $7M to tap the conversational AI opportunity across Asia  

SleekFlow, a social commerce platform based in Singapore and Hong Kong, secured $7 million in funding to enhance its AI-driven customer engagement tools and expand into Southeast Asia, the Middle East, and Europe. The startup offers omnichannel marketing, automation, and CRM integrations, targeting businesses in insurance, healthcare, telecom, and retail sectors. TECHCRUNCH