Sakhanda Wire
NVDA $155.46 -0.44% MSFT $506.24 -2.27% GOOGL $280.20 -1.24% META $618.87 -1.20% AMZN $257.89 +0.05%

Latest news

Older stories

The Verge

Researchers question OpenAI’s selective sharing of research results

OpenAI announced breakthroughs on Monday in mathematics, computer science, and physics, while also describing what it called an evolving approach to communicating its findings. The announcement came as researchers who spoke to The Verge expressed growing concern about how the company controls access to some results.

TechCrunch

Disruptive launches $10 billion fund for late-stage companies

Dallas-based investment firm Disruptive has announced a $10 billion fund aimed at backing at least 10 late-stage companies. The firm said individual investments could reach $1 billion, with the vehicle focused on capital-intensive areas including artificial intelligence and defense. It is among the largest funds raised so far this year.

Ars Technica

Natural-language prompts enter protein and DNA design tools

Protein-design software is beginning to accept natural-language descriptions of the function a desired protein should perform. A large collaboration of researchers based in China has developed a tool that turns such prompts into protein designs, reducing the need for users to specify every structural detail themselves.

The Verge

VisionQuest turns AI fears into a surreal story about humanity

Marvel’s VisionQuest follows a superpowered robot trying to understand humanity. He becomes lost in elaborate digital daydreams where he and other artificial intelligence-powered beings share a home. The series is highly fantastical, but it uses this setup to explore questions about machine intelligence, identity and coexistence.

The Decoder

Google DeepMind unveils Genie 4, a real-time interactive world model

Google DeepMind has introduced Genie 4, an experimental world model that calculates interactive 3D environments frame by frame rather than producing finished videos. Users can move through scenes with a keyboard and influence events using text prompts. For the first time in the Genie series, DeepMind says the system generates matching background audio directly, including environmental sounds, and can produce dialogue between characters that respond to users and events.

Hugging Face

LFM3 brings specialized multimodal models onto one backbone

LiquidAI has announced LFM3, a new generation of Liquid Foundation Models designed as a base for specialized applications. The family covers text, vision, audio, omni and decision tasks, with six models built from the LFM3-1.7B-Base backbone. Developers can fine-tune each model on a single GPU and deploy it on the device where an application runs.

TechCrunch

AI labs challenge a foundational assumption about computational complexity

OpenAI has released work addressing three major mathematical problems, including a proposed refutation of the Exponential Time Hypothesis (ETH), a long-standing assumption in computer science. ETH holds that certain difficult logic problems require exponentially more time as the number of variables increases. OpenAI and Anthropic say their models produced algorithms that solve these problems in sub-exponential time, which would contradict ETH if the work is confirmed.

The Verge

Anthropic Designs Governance Plan Ahead of Impending IPO

Anthropic is preparing for an impending initial public offering while seeking to preserve its stated mission of developing artificial intelligence safely and for the benefit of humanity. The company aims to access the substantial funding available through public markets without allowing that transition to alter its priorities.

TechCrunch

Google DeepMind previews Genie 4 with audio and real-time world editing

Google DeepMind has released Genie 4 as a research preview, presenting it as a more photorealistic and interactive world model than its predecessor. Users can enter text instructions while a simulation is running to alter the weather, camera angle, people, objects, or other aspects of the scene. The system also produces audio synchronized with the simulated environment, including music, ambient sounds, and interactive dialogue.

Hugging Face

Study finds LLM judgments of research novelty are highly unstable

A new study from Ai2 examines how reliably large language models can determine whether scientific research ideas are novel. The researchers released the Old Ideas, Novel Problems benchmark, data and code, along with a pipeline for creating additional evaluation sets. The benchmark tests both absolute novelty ratings and comparisons between two ideas.

DeepMind

Genie 4 brings real-time audio and interactive world simulation

On October 27, 2026, DeepMind announced Genie 4, a world model designed to make real-time interactive simulations more realistic. The company says the system is natively multimodal, generating music-like audio, environmental sounds and interactive dialogue alongside the visual world. Gemini Omni supplies knowledge of real-world appearance and behavior, while other Gemini models provide reasoning intended to preserve consistency and continuity.