AI Intelligence — AI-assisted research, 48+ sources, human-curated
Demis Hassabis formed a covert ~20-researcher team within DeepMind to develop AI-powered high-frequency trading algorithms, reportedly targeting rival Renaissance Technologies. Google leadership disapproved, leading to the project's quiet termination.
Google Research has published a technical paper introducing TurboQuant, a new AI model quantization method that reportedly reduces memory usage by 6x and could cut AI inference costs by 80%. The research suggests significant implications for AI infrastructure economics and hardware investment strategies.
A new benchmark claim suggests DeepSeek-R1 has achieved 78.9% on the OS-World agentic coding benchmark, reportedly outperforming GPT-5.4 while operating at one-tenth the cost. If verified, this would represent a significant leap in cost-performance for AI coding agents.
Researchers from Stanford and Harvard have published a notable AI paper focusing on mechanistic interpretability and AI safety, with implications for understanding and securing advanced AI systems.
OpenAI has closed a $122 billion funding round at an $852 billion valuation, led by Amazon, Nvidia, and SoftBank. The company disclosed $2 billion in monthly revenue, 900M+ weekly users, and is positioning for a public offering.
Demis Hassabis formed a covert ~20-researcher team within DeepMind to develop AI-powered high-frequency trading algorithms, reportedly targeting rival Renaissance Technologies. Google leadership disapproved, leading to the project's quiet termination.
Apr 2, 2026
AI-generated daily podcast
DeepSeek says it just beat GPT-5.4 on OS-World for one-tenth the cost, and that alone would be enough to start a bar fight. But the weirder story is that memory, energy, and even prompt files are turning into the real battleground — while Google quietly pokes at quantum Bitcoin and trading desks like it’s rearranging the furniture in a haunted house.
Google Research has published a technical paper introducing TurboQuant, a new AI model quantization method that reportedly reduces memory usage by 6x...
Microsoft has unveiled a significant upgrade to its Copilot AI assistant, enabling users to employ multiple AI models simultaneously within a single w...
A new benchmark claim suggests DeepSeek-R1 has achieved 78.9% on the OS-World agentic coding benchmark, reportedly outperforming GPT-5.4 while operati...
Researchers from Stanford and Harvard have published a notable AI paper focusing on mechanistic interpretability and AI safety, with implications for...
OpenAI has closed a $122 billion funding round at an $852 billion valuation, led by Amazon, Nvidia, and SoftBank. The company disclosed $2 billion in...
Researchers introduced MemFactory, a unified framework treating agent memory as a trainable component. It supports multiple memory paradigms and shows...
Google researchers have compiled Shor's algorithm to solve Bitcoin's 256-bit elliptic curve problem with ~1.2k logical qubits, translating to <500k ph...
Superintelligence is launching a new Sunday edition called 'Intelligence from the Community,' opening its platform to external contributors. Selected...
Google has open-sourced TimesFM, a foundation model for time series forecasting trained on 100 billion real-world time points. It requires no dataset-...
Apple's senior vice president of services, Eddy Cue, will appear live on the TBPN podcast today at 12:10 PM PT. The interview is part of Apple's 50th-...
Renewable energy reached 49.4% of global electricity capacity in 2025, adding 692 GW in a single year. Solar contributed 511 GW, becoming the primary...
MCP's Role in Shaping Future Agentic AI Standards
Superintelligence is launching a new Sunday edition called 'Intelligence from the Community' that will feature original research, sharp insights, and theoretical ideas from the AI community, offering contributors access to 225,000 engaged readers.
Zai has launched GLM-5V-Turbo, a multimodal model that processes images, videos, and documents to generate runnable code from visual inputs, claiming top benchmark performance in design reconstruction and visual coding tasks.
AI researcher Ethan Mollick observes that despite AI's ability to generate any image or video for free, human creativity remains a bottleneck, as evidenced by April Fools' jokes not improving in quality.
Demis Hassabis reportedly assembled a secret team at DeepMind to develop AI-powered high-frequency trading algorithms to rival Renaissance Technologies, but Google disapproved and the project was disbanded.
A new AI model reportedly achieves 78.9% on OS-world benchmark while outperforming GPT-5.4 at one-tenth the cost, representing a significant efficiency breakthrough in large language models.
Newsletters summarize yesterday. We build a living knowledge graph and make predictions you can verify.
3,111 entities with typed relationships — structured intelligence you can query via API.
Explore entities →Falsifiable predictions with confidence scores, auto-verified. 77.8% on 98 resolved.
See predictions →AI-assisted pipeline updates every 2–6 hours — scanning, analyzing, and publishing with editorial oversight at every stage.
How it works →