Black Forest Labs launches FLUX.1 text-to-image model suite, attracting $31 million in seed funding. FLUX.1 offers closed-source [pro] and open-source [dev] versions, featuring 12 billion parameters and innovative hybrid architecture. Experts recognize FLUX.1's output quality surpassing closed-source models, driving innovation in graphic design and visualization. The release sets new standards in responsible AI technology development, introducing "flow matching" method and enhancing visual quali

Microsoft earns nearly $20 million monthly from TikTok through OpenAI access, a significant part of its cloud revenue. TikTok aims to reduce dependency by developing its own language model, despite ByteDance's past misuse of OpenAI tech. Microsoft, as OpenAI's exclusive cloud provider, invests heavily in AI infrastructure, with Azure revenue showing strong growth potential.

Runway introduces Gen-3 Alpha Turbo, 7x faster than the original model. Pricing will decrease while maintaining quality. Updates for mobile app in progress. Potential training dataset concerns raised by critics.

The 2024 JD Cloud Technology Summit showcased JD Cloud's AI breakthroughs, launching cutting-edge products like Enterprise Large-scale Model Service, Yanshi AI Development Computing Platform 2.0, and Large-scale Model Security and Trust Platform. These innovations empower industries and drive intelligent transformation.

Baidu introduces self-reasoning AI framework for accurate decision-making, overcoming hallucination issue in language models. It enhances transparency, reliability, and accountability, paving the way for trustworthy AI systems in various industries.

Google expands Gemma 2 series with powerful 2B model, surpassing large-scale AI models. Optimized for NVIDIA's TensorRT-LLM library, Gemma 2B is efficient across hardware environments. ShieldGemma suite filters harmful content; Gemma Scope toolkit promotes transparent AI development.

Google introduces "Circle to Search" to verify image authenticity, displaying metadata & source info. Useful for verifying political images on social media, reducing spread of misleading info. Rolling out on Pixel series & Samsung devices to combat false info & enhance user experience, paving the way for a more authentic digital environment.

This week, Neura unveiled their humanoid robot 4NE-1, showcasing its flexibility in tasks like ironing and box moving. Nvidia also introduced new humanoid robot development tools. Neura collaborates with industry giants like Nvidia, Boston Dynamics, and ByteDance. Neura's CEO confirmed 4NE-1's capabilities to be displayed in July. The video highlights 4NE-1's efficiency, but it's essential to remain rational about its operational level. Neura enters the humanoid robot field and partners with Nvi

Meta's Q2 earnings show strong revenue and profit growth, with bold AI investments for future returns. Zuckerberg anticipates AI assistant dominance and innovative products like Llama 4 and AI glasses. Quest 3 surpasses sales expectations, with a cheaper headset on the way. Threads sees user growth, while Facebook gains more young adult users.

Google is launching innovative security measures to remove deepfake content from search results, protecting users from explicit images. Updates reduce explicit images by over 70% and prioritize non-explicit content. Google aims to purify the online environment while safeguarding user privacy.

The latest iPhone 15 Pro AI update introduces language improvements and innovative features like text interaction. AI elements are also found in Mail app with new Writing Tools. Apple Intelligence streamlines tasks like creating summaries and tables. The potential for Siri to extract information and perform tasks within apps is promising for future updates.

Midjourney released v6.1 with significant upgrades in character depiction and texture mapping. The update enhances image quality, text accuracy, speed, and personalization, setting the stage for the platform's future development towards v7.

Google has strengthened search function to combat deepfakes, removing explicit content and improving user control. Updates reduce exposure of harmful content by over 70% and prioritize high-quality search results. "About this image" feature added to Google Lens for easy verification.

Canva announces acquisition of Leonardo.ai for visual AI toolkit expansion. Integration of customizable text-to-image and video tech. Co-founder confirms independent platform development. Enhances competitiveness against Adobe. Focus on data privacy and transparency. Expect advanced AI features for users.

OpenAI releases GPT-4o Long Output language model with 64,000 token output capacity, addressing customer demand for longer responses. Priced at $6 per million input tokens and $18 per million output tokens, aiming for wider developer accessibility.

Google added a new feature to Chrome 128 Beta, allowing users to search selected areas on web pages with Google Lens, improving information acquisition efficiency and browsing experience. This feature is expected to be officially launched soon.

Shengshu Technology launched the Vidu AI video model for global creators, offering text-to-video features, diverse styles, and image-to-video capabilities. The tool provides efficient video creation and advanced options, aiming to revolutionize the field of video creation.

AMD reported strong Q2 2024 financials, with data center sales accounting for nearly half revenue. This growth is mainly driven by Instinct MI300 accelerator. AMD plans to launch new AI chips annually, with MI325X in Q4 2024. Although behind Nvidia in data center sales, AMD also saw growth in PC CPU and GPU business. Over 100 platforms will use Ryzen AI 300 "Strix Point" chip for laptops.

Meta shifts focus from celebrity AI chatbots to AI Studio for creators; users can now create their own AI chatbots. The move comes after shutting down celebrity chatbots and launching AI Studio tool, aiming to provide a platform for diverse AI creation.

OpenAI releases advanced voice mode for ChatGPT Plus subscribers, delayed for safety concerns. Improved response to interruptions with scheduled release in fall 2023, featuring preset voices to avoid mimicry concerns.