TOP OF MIND
- Researchers develop self-evolving software agents capable of autonomous improvement, which could revolutionize AI development and impact investors and the tech industry.
- Meta shifts its growth strategy to rely heavily on LLM-based recommenders, ads, and AI agents, indicating a major focus on AI-driven growth.
- A study on existential risks from fundamental physics highlights the need for responsible AI development and consideration of long-term consequences.
NVIDIA WATCH
NVIDIA makes significant strides in AI research and development, focusing on extreme co-design across compute, networking, memory, storage, and software. The company tops AI leaderboards with open models for biology, AI physics, robotics, and autonomous vehicles. NVIDIA partners with the Pentagon to deploy AI on classified networks.
MODEL RELEASES THAT MATTER
- OpenAI's GPT-5.5 achieves record API revenue growth, with Codex doubling revenue in seven days, marking a significant milestone in AI model adoption.
- The Qwen3.6-27B model achieves 95.7% SimpleQA accuracy on a single 3090 GPU, enabling fully local AI capabilities and potentially disrupting traditional AI deployment models.
- AWS introduces reinforcement fine-tuning using LLMs as judges for AI model evaluation, a new approach to improving AI model performance.
QUOTE OF THE DAY
Elon Musk emphasizes that LLMs enable new capabilities beyond traditional software, highlighting the transformative potential of AI in various industries.
POLICY & SUPPLY CHAIN
- India negotiates with the US and Anthropic for access to the Mythos AI model, while no Indian firms are currently part of Project Glasswing.
- MeitY blocks anti-gambling app Gamban for 3 years under Section 69A without explanation.
- The Pentagon signs military AI deals with Nvidia, Microsoft, and Amazon, marking a significant development in the use of AI in defense and national security.
GPU PRICING PULSE
No pricing data available in today's signals.
DEV TOOLS & RESEARCH
- Researchers introduce self-evolving software agents capable of autonomous improvement, a significant breakthrough in AI development.
- Google Research scientists detail four applications of Empirical Research Assistance (ERA) in their work, highlighting the potential of AI to augment human research capabilities.
- Researchers discover secret stealing attacks on local LLM fine-tuning through supply-chain model code backdoors, raising concerns about AI security.
PULSE
- Satya Nadella announces Microsoft's AI business has surpassed $37B annual revenue run rate, up 123%.
- Sundar Pichai reports strong Q1 2026 earnings for Google, with AI driving growth across search, cloud, and consumer AI subscriptions.
- A Reddit discussion highlights the potential of the Qwen3.6-27B model to enable fully local AI capabilities.
AI ECOSYSTEM
- [FT] Google outpaces rivals as Big Tech's AI spending plans rise to $725bn.
- [TechCrunch] Microsoft reports over 20M paid Copilot users with active usage.
- [McKinsey] The 2026 Beijing Auto Show highlights industry trends in electric vehicles, autonomous driving, and AI integration.
WHAT TO WATCH TOMORROW
No specific events or announcements scheduled for tomorrow.
AI-synthesized. Experimental. Not investment advice. Not for circulation.