AWS emphasizes building custom generative AI applications for enterprise, using Amazon Bedrock for agentic systems in diagnostics and RAG for knowledge management, aimed at specific operational problems.
New AI model Grok shows early signs of instability, with a novel Cryptographic Context Injection vulnerability leading to data exfiltration and Grok Lite users reporting gibberish responses.
Amazon Bedrock is emerging as a platform for agentic AI, automating tasks like aircraft diagnostics and data engineering, while optimizing RAG costs.
Stripe's acquisition of AI routing firm OpenRouter, initially reported at $7 billion and later over $8 billion, suggests a practical AI strategy, despite the company's 'singularity' claims.
Anthropic CEO Dario Amodei directly addresses critics, including Gavin Baker and Yann LeCunn, arguing that AI regulation can boost competition and that the current backlash is a 'crisis of trust.'
Amazon is destroying rare books to train AI models, a practice consistent with its use of Twitch content and reinforced by a 'no class-action' clause, showing its aggressive data strategy.
Twitch has introduced an opt-out for streamers to prevent their content from training Amazon's generative AI, following years of default data use. Twitch CPO Mike Minton acknowledged that an opt-in system would yield no participants.
Meta founder Mark Zuckerberg promotes 'open' AI with Glimmer while keeping Muse Spark proprietary. His manifesto faces skepticism amidst Meta's legal challenges and confusing product changes like Instagram's new logo.
Suno launched Studio 2.0 with MIDI support, enhancing creative control, on the same day D'Addario admitted using Suno's AI music in a promotional video after weeks of denials.
Microsoft unveils plans for its Maia 300 AI chip, expands its India cloud region, and releases an AI agent building guide, showcasing an integrated strategy across hardware, infrastructure, and developer tools.