9th October - AI News Daily - Google's Gemini 2.5 Unleashes Browser Automation, Reshaping Agent Capabilities

Send us a text

🌍 INAI • The Open AI Hub

The Intelligence Atlas → the world’s most comprehensive, open hub of AI knowledge. 2 Million+ tools, models, agents, tutorials & daily news—free for all, updated every day.

https://github.com/inai-sandy/inAI-wiki

TOP HIGHLIGHTS

Google's Gemini 2.5 introduces "computer use" capabilities for browser automation, bringing agent automation to the mainstreamAMD secures multi-billion GPU deal with OpenAI while Nvidia tightens direct sales, intensifying AI compute competitionSecurity concerns emerge with first malicious MCP server discovery and Figma MCP vulnerabilityCoreWeave launches Serverless RL with Weights & Biases integration to simplify agent trainingDisney and Universal sue Midjourney over character imagery, escalating copyright debates

NEW TOOLS & FRAMEWORKS

Microsoft unifies AutoGen and Semantic Kernel into enterprise-ready Agent FrameworkAnthropic releases Petri for open-source LLM auditingGoogle's Opal no-code app builder expands to 15 countriesStripe adds model pricing and usage tracking APIsPython 3.14 stabilizes GIL-free interpreter with Pydantic 2.12 support

LLM INNOVATIONS

Ling-1T debuts trillion-parameter open-source reasonerSamsung's 7M-parameter Tiny Recursive Model outperforms larger systemsAI21's Jamba Reasoning 3B offers efficient reasoning trade-offsAlibaba releases Qwen3 Omni multimodal model and Qwen Image EditLiquidAI demonstrates on-device reasoning for iPhone 17 Pro

RESEARCH HIGHLIGHTS

Drax achieves SOTA speech recognition with discrete flow matchingModernVBERT outperforms larger models through architecture innovationMulti-vector embeddings improve retrieval precisionCAIS updates "Humanity's Last Exam" to rolling benchmarkVChain introduces chain-of-visual-thought for video reasoningResearch shows quantization resilience must be built into training

INDUSTRY & POLICY DEVELOPMENTS

USPTO pilots AI-assisted prior-art discovery for patent applicationsGoogle faces DOJ scrutiny over Gemini integration in core servicesHidden Unicode payload attacks affect some LLMs, including Gemini-class models

PRACTICAL RESOURCES

Step-by-step RAG implementation guide for beginnersGuide on when to parse vs. extract in document workflowsStrategies for Sora 2 guardrails and watermarkingPrompt optimization techniques for agent reliabilityPrivacy best practices for biometric data handling

DEMOS & APPLICATIONS

Intercom showcases LangGraph powering Fin_ai customer supportPika's Predictive Video enables prompt-to-clip creationSora-powered "viral video recreator" teasedSeedream mobile agent enables on-device image generationCristiano Ronaldo reportedly used Perplexity AI for speech preparation

THOUGHT-PROVOKING DISCUSSIONS

JEPAs may bridge generative and contrastive learningQuality over quantity emphasized for RL training signalsStudies show sycophantic AI undermines relationship repairLLM checks identify 80M+ inconsistent Wikipedia factsIndustry consolidation raises concerns about AI infrastructure accessSora's upside-down exploit highlights evaluation gaps

Support the show