All Updates

Page 819 of 1664

April 28, 2026

๐Ÿ‡ฌ๐Ÿ‡ง
The Register - AI/MLโ€ข89d ago

GitHub Copilot Ends Unlimited AI Billing

Microsoft's GitHub is shifting GitHub Copilot from an all-you-can-eat model to metered AI billing due to a cost crisis. The company acknowledges it cannot sustain unlimited AI access like Red Lobster's Endless Shrimp promotion. This change addresses unsustainable flat-rate consumption.

#metered-billing#ai-costs#developer-tools
๐Ÿงง
Qwen (GitHub Releases: qwen-code)โ€ข89d ago

Qwen Code v0.15.2 Nightly: CLI & VSCode Boosts

Qwen Code's v0.15.2-nightly release adds Catalan language support and CLI features like API preconnect for reduced latency and slash command hints. It includes VSCode companion fixes, LLM-generated tool-call summaries, and core agent controls. New contributors and various bug fixes improve stability.

#cli-tools#vscode-extension#agent-controls
๐Ÿค—
Hugging Face Blogโ€ข89d ago

Physics-Informed AI for Adaptive Ultrasound

Hugging Face blog announces NV-Raw2Insights-US, a physics-informed AI model for adaptive ultrasound imaging. It processes raw ultrasound data to generate high-quality images. This advances medical imaging with AI integration on an open platform.

#ultrasound#medical-ai#physics-informed
๐Ÿ“Š
Bloomberg Technologyโ€ข89d ago

Advantest Stock Drops on Capacity Crunch

Advantest Corp.'s shares fell as much as 6.9% after a weak outlook. The AI chip tester maker cited ongoing capacity constraints. This highlights supply challenges in AI hardware testing.

#chip-testing#supply-chain#semiconductors
๐Ÿ’ฐ
้’›ๅช’ไฝ“โ€ข89d ago

OpenAI Integrates Codex into GPT-5.5, Plans AI Phone

Microsoft and OpenAI end revenue sharing for non-exclusive licensing. Anthropic's Claude completes 186 autonomous transactions, causing eBay stock to drop 4.5%. OpenAI terminates standalone Codex, merges it into GPT-5.5, and plans self-developed AI phone hardware by 2028.

#ai-agents#partnerships#ai-hardware
๐Ÿ“Š
Bloomberg Technologyโ€ข89d ago

Nine Jurors Seated in Musk-Altman Trial

Nine jurors were seated in the high-profile Elon Musk vs. Sam Altman trial. The case involves OpenAI and is set for three weeks in federal court. Selected from San Francisco Bay Area residents.

#lawsuit#trial#ai-governance
๐Ÿ•๏ธ
ๆžๅฎขๅ…ฌๅ›ญโ€ข89d ago

HappyHorse 1.0 Pricing Drops to 0.44 Yuan/720P Sec

Alibaba launches HappyHorse 1.0 gray test on official site and Aliyun, with 720P video at 0.9 yuan/sec (0.44 discounted). Pros access via website with free credits; integrates with agents like Ali Wukong. Targets ads, ecom, short drama with full creation tools.

#ai-video-pricing#enterprise-access#agent-integration
๐Ÿ“‹
TestingCatalogโ€ข89d ago

OpenAI Enables Multi-Cloud Model Hosting

OpenAI and Microsoft updated their partnership to let OpenAI host models on any cloud provider, including AWS. Azure remains the priority cloud for optimized performance. This expands deployment flexibility for OpenAI products.

#partnership-update#multi-cloud#deployment
๐Ÿฏ
่™Žๅ—…โ€ข89d ago

Ford CEO Praises Chinese EV AI

Ford CEO Jim Farley praises Xiaomi SU7's AI features and BYD's integration after personal testing, but opposes US imports amid tariffs. US automakers retreat from EVs, writing off billions while China advances in software-defined vehicles.

#ev-ai#china-auto#vertical-integration
๐Ÿฏ
่™Žๅ—…โ€ข89d ago

TSMC Delays High-NA EUV to 2029

TSMC pauses procurement of costly ASML High-NA EUV until 2029 A13 node, optimizing existing EUV instead to protect profits. Decision pressures ASML stock but affirms TSMC's cost control amid AI demand uncertainty.

#lithography#chipmaking#semicon-delay
๐Ÿฏ
่™Žๅ—…โ€ข89d ago

Meta Launches Muse Spark Post-Manus Fail

Meta's $2B Manus AI agent acquisition blocked by China, forcing launch of internal Muse Spark as closed commercial model. Shifts from Llama open-source to monetized agents, boosting stock 9.5%. Proves self-reliance under M&A risk.

#ai-agent#acquisition-fail#meta-strategy
๐Ÿ“‹
TestingCatalogโ€ข89d ago

Copilot Adopts Usage-Based Billing in 2026

GitHub Copilot shifts to usage-based billing for all plans starting June 1, 2026, based on AI credits and tokens. Admins can set spending limits. Bill previews available in May 2026.

#usage-billing#ai-credits#cost-limits
๐Ÿ“Š
Bloomberg Technologyโ€ข89d ago

Dynatrace Surges on Starboard's AI Push

Dynatrace shares rose over 6% in late trading following reports that activist investor Starboard Value LP has taken a stake. Starboard is pressing the company to better capitalize on the artificial intelligence shift. This could signal upcoming strategic changes in the observability platform.

#activist-investor#stock-rally#ai-strategy
๐Ÿ”ฅ
36ๆฐชโ€ข89d ago

DeepSeek Slashes Input Cache Prices 90%

DeepSeek reduced input cache hit prices across all API services to 1/10th starting April 26. Pro models add 2.5-fold discount until May 5, 2026. V4-Pro now costs 0.025 CNY/million tokens, V4-Flash 0.02 CNY/million tokens.

#api-pricing#inference-costs#llm-api
๐Ÿ“ฐ
The Vergeโ€ข89d ago

Google Tests AI Chatbot Search on YouTube

Google is testing a conversational AI search experience on YouTube, similar to AI Mode, that pulls longform videos, YouTube Shorts, and text results. The experiment is available to YouTube Premium subscribers in the US who are 18 or older, featuring an 'Ask YouTube' button with prompts like funny clips or summaries. It aims to make searching feel more like a conversation.

#conversational-ai#multimodal-search#video-platform
๐Ÿ”ฅ
36ๆฐชโ€ข89d ago

Xingji Interactive Raises Angel Funding for AI IP

Xingji Interactive, an AI-driven entertainment firm, secured tens of millions RMB in angel funding from Zhengda Group and others. Funds target team growth to 120, production ramp-up, and platform development. Focus on Phoenix AI for full-stack video workflows and IP like digital humans and manhua dramas.

#ai-funding#ai-ip#generative-video
๐Ÿค–
OpenAI Newsโ€ข89d ago

OpenAI's ChatGPT Safety Commitment

OpenAI details its dedication to community safety in ChatGPT via multiple layers. These include model safeguards, misuse detection, policy enforcement, and partnerships with safety experts. The initiative ensures safer user interactions.

#community-safety#model-safeguards#policy-enforcement
๐Ÿค–
OpenAI Newsโ€ข89d ago

OpenAI Models Launch on AWS

OpenAI GPT models, Codex, and Managed Agents are now available on AWS. This enables enterprises to build secure AI applications directly in their AWS environments. It bridges OpenAI's capabilities with AWS infrastructure for compliant deployments.

#cloud-integration#enterprise-ai#model-deployment
๐ŸŽ
Apple Machine Learningโ€ข89d ago

LaDiR Boosts LLM Reasoning via Latent Diffusion

Apple's LaDiR framework integrates latent diffusion models with LLMs to enhance chain-of-thought reasoning. It overcomes autoregressive decoding limits by enabling holistic token refinement in a continuous latent space. This promotes efficient exploration of diverse reasoning solutions.

#latent-diffusion#chain-of-thought#text-reasoning
๐ŸฆŠ
GitLab Blogโ€ข89d ago

GitLab Deepens Claude Integration for Governed AI

GitLab has deepened integration with Anthropic's Claude models, making them the default in the GitLab Duo Agent Platform for code generation, review, agentic chat, and vulnerability resolution. This ensures AI operates within enterprise governance, compliance, and audit trails in the SDLC. Enterprises can access Claude via Google Cloud Vertex AI, AWS Bedrock, or the Claude Marketplace.

#governed-ai#agentic-workflows#devsecops
Page 819 of 1664