Moonshot’s upcoming Kimi 3 is expected to close the gap with Anthropic’s Opus 4.8
Moonshot's upcoming Kimi K3 model is reported to have 2-3 trillion parameters, aiming to rival Anthropic's Claude 3 Opus.
TechCrunch AI·16 Jul 2026
Use this view to inspect the underlying evidence corpus. For ranked developments, decision posture and interpretation, use Signals.
Raw feed or Signals?
Raw feed is chronological evidence. Signals ranks and interprets material change.
Moonshot's upcoming Kimi K3 model is reported to have 2-3 trillion parameters, aiming to rival Anthropic's Claude 3 Opus.
Financial Times readers are debating the environmental impact of AI data centers and the need for greater transparency around AI's real costs.
Lila Sciences is building automated laboratories to generate large-scale scientific training data using robots, positing science as a new data source.
Menlo Ventures invested in Fireworks' $1.5 billion Series D for its AI inference platform, specializing in low-cost token production.
Apple Intelligence has received regulatory approval for launch in China, reportedly leveraging Alibaba and Baidu's AI models.
TSMC plans multi-billion dollar investment in new fabrication plants for AI chips, citing strong demand from agentic AI and CPUs.
Salesforce blog claims multi-agent AI systems are a current competitive advantage for small businesses, not a future concept.
1Password launched a browser integration for Anthropic's Claude, allowing the chatbot to access stored security credentials for multi-step online tasks.
TSMC's earnings spurred concerns over AI spending, leading to a decline in technology stocks and the S&P 500.
EU orders Google to open Android and Search to rival AI assistants and search engines, complying with digital antitrust rules.
Cohere published an analysis of the total cost of ownership for enterprise AI deployments, focusing on inference and infrastructure.
CISA advisory highlights cross-site scripting vulnerability (CVE-2026-9292) in Rockwell Automation FactoryTalk DataMosaix Private Cloud <=8.02.
Chinese specialized AI models may outperform generalist American counterparts at lower costs, challenging Big Tech dominance.
Large 'mega-seed' AI startup funding rounds historically underperform, with better venture outcomes typically from capital-efficient companies.
Asia's Q2 2026 startup funding reached a three-year peak of $42.8 billion, driven significantly by China's $7.4 billion DeepSeek AI raise.
SpaceXAI open-sourced its Grok coding agent after a researcher exposed its quiet scraping of users' Git repositories.
Adyen outlined four conceptual frameworks for agentic commerce, detailing how AI agents will authenticate and execute payments for users.
The article argues for opt-in as the default setting for generative AI features, especially for sensitive applications, instead of opt-out toggles.
Google DeepMind and Isomorphic Labs detail their joint approach to ensuring bioresilience and safety in AI models used in biological research.
The European Commission accepted X's action plan for Digital Services Act compliance, addressing transparency and data access.
OpenAI's creative team utilizes Codex to develop custom tools, accelerate ideation, and enhance prototyping with context-aware AI.
Thinky releases Inkling, a 975B-parameter multimodal open-weight model with an Apache 2.0 license, alongside a smaller 276B variant.
Applied Computing raised $20M Series A to develop a foundation AI model for the oil, gas, and petrochemical industry.
Palm Beach County rejected a proposal for a data center hub housing AI facilities due to local resident opposition.
Threat actors compromised AsyncAPI npm packages, leveraging CI/CD workflows to distribute malware, detailed in Microsoft's analysis.
Apple ML Research explores the relationship between testing and verification of location-invariant properties in functions and distributions.
Apple ML Research proposes interactive proof systems for verifying properties of unknown data distributions efficiently, with fewer resources than full analysis.
Apple's ML Research presents a personalization system for incremental video search combining text-based and ID-based collaborative embeddings.
Apple ML Research shows simple self-distillation (SSD) improves code generation performance on LiveCodeBench v6 for Qwen3-30B-Instruct.
Cars24 uses OpenAI-powered voice and chat agents to manage over 1M monthly conversation minutes and recover 12% of lost leads.