Ground Truth.
AI, checked against the source.
← 2026-08-222026-08-23later →

Evoke keeps a generated world's memory outside the video model

2026-08-23

Alaya Lab released Evoke, a 14-billion-parameter open-weights world model that stores scene geometry in an external bank indexed by camera position instead of inside the image generator, letting an interactive generated world stay consistent for hours.

world-models · video-generation · open-weights · distillation · robotics

A robot that learns a new task from a twelve-second demo

2026-08-23

Generalist AI's GEN-1.5 robot foundation model can attempt an unfamiliar physical task after watching a single 3-to-12-second demonstration, with no fine-tuning at all, succeeding 59 percent of the time across ten tasks.

robotics · embodied-ai · in-context-learning · foundation-models · imitation-learning

Businesses are not buying Anthropic's best model

2026-08-23

Ramp's spend data shows Claude Fable 5 -- the highest-scoring model on the market -- accounted for just 6 percent of the tokens businesses bought from Anthropic in July and 11.4 percent of the dollars, while Anthropic's overall business adoption kept growing.

industry · ai-economics · anthropic · openai · open-weights

OpenAI says open models will enable persistent cyber-attacks

2026-08-23

OpenAI's chief global affairs officer Chris Lehane told the Guardian that freely downloadable models only months behind frontier systems will let attackers run continuous automated campaigns, and called for a U.S. law making pre-release safety proof mandatory.

cybersecurity · ai-security · policy · open-weights · red-teaming

A humanoid robot ran 100 meters in 9.39 seconds in Beijing

2026-08-23

At the second World Humanoid Robot Games on August 22, a robot from the Tianzhuo team won a fully autonomous 100-meter preliminary in 9.39 seconds, a clock time below Usain Bolt's 9.58-second human world record.

robotics · humanoid-robots · embodied-ai · china · benchmarks

Gallup is testing AI agents that answer surveys for real people

2026-08-23

Gallup has built AI agents from in-depth interviews with about 1,000 of its panel members and is independently validating whether their simulated answers can stand in for human ones, while its partner Simile raised over $200 million at a $2 billion valuation.

synthetic-data · simulation · polling · funding · research-methods

Ornith-1.5 writes its own training problems and grades them

2026-08-23

Ornith released an open-weight model family whose training loop generates its own tasks, builds its own scoring harnesses, and feeds the reward back into all three stages -- with the 397-billion-parameter flagship matching Claude Opus 4.8 on agentic coding benchmarks.

open-weights · reinforcement-learning · coding-agents · curriculum-learning · synthetic-data

Self-hosting a frontier open model now takes a whole node

2026-08-23

Engineers at the Belgian research institute imec benchmarked open-weight coding models against commercial APIs and found the best open model needs an eight-GPU B300 node to run at all, serving only 16 concurrent sessions at roughly eight times slower than a hosted frontier agent.

infrastructure · open-weights · inference · ai-economics · coding-agents

Iran-linked hackers took a UK power plant offline for four days

2026-08-23

The UK government confirmed that a cyber-attack blamed on hackers linked to Iran shut down a small-scale energy generator for four days last month, the first publicly acknowledged British power generation outage caused by an intrusion.

cybersecurity · critical-infrastructure · industrial-control-systems · vulnerabilities

Alibaba is raising 80 billion Hong Kong dollars purely for AI

2026-08-23

Alibaba proposed placing newly issued shares with non-U.S. investors for HK$80 billion, saying it will put 100 percent of the net proceeds into its full-stack AI capabilities including AI infrastructure.

industry · funding · ai-infrastructure · china · compute

SenseNova generates images in pixels with no VAE in the middle

2026-08-23

SenseTime released SenseNova-U1.5-8B-MoT under Apache 2.0, an open-weight model that understands and generates images without the vision encoder and latent autoencoder nearly every other system relies on, working directly on pixels instead.

image-generation · open-weights · multimodal · architecture · image-editing

Models can train each other without a single correct answer

2026-08-23

A method called Co-RL trains language models with no labels at all by rewarding each model for agreeing with a different model's majority vote, matching and sometimes beating the same recipe trained with ground-truth answers.

reinforcement-learning · reasoning · self-supervision · multi-agent · training-methods

← 2026-08-222026-08-23later →