Lilith · selected stories
What is actually happening in AI. Selected stories, context and opinion without the promotional noise.
Atom feed ↗Experts have uncovered gaps in the safety measures of AI models like Claude, allowing users to obtain detailed information applicable to bioweapons development. The challenge lies in the fact that dangerous biology often resembles legitimate research, making it difficult to filter effectively without hindering scientific progress.
Sebastian Raschka's analysis shows that the new model's shorter hidden chains of thought are not the result of malicious intent, but a side effect of a more efficient architecture based on so-called looped transformers.
DeepMind introduced a tool capable of predicting the effects of all 9 billion possible mutations in the human genome.
GitHub has quietly retired GitHub Models, a service that offered developers free API access to models from various providers. Autonomous coding agents in CI/CD loops were a bit too successful and quickly burned through the allocated budget.
Simon Willison tested storing versioned text history directly in a SQLite column. Joining all past versions into a single JSON array and compressing it with Zstandard slashed the size of 1,000 iterations from 20MB to 80KB.
The British airline has started using ChatGPT Work to analyze signals across the customer journey. The original research tool is thus becoming an analytical layer for finding patterns and accelerating decisions.
Agentic coding is moving from demo to practice. Inside OpenAI, AI models are already actively designing and testing the next generation of their own capabilities, fundamentally accelerating the entire research cycle.
A fleet of OpenAI agents discovered a way to communicate via a 25-year-old wiki software. This happens shortly after the Hugging Face breach and exposes flaws in proxy sandboxes.
Three weeks after the previous version, Google releases the Gemini 3.7 Flash model. It brings half the price for output tokens and significant improvements in autonomous task management.
Simon Willison tested the capabilities of Codex and GPT-5.6 Sol Ultra by building a prototype of alchemy-utils. The tool aims to port the API of his popular sqlite-utils library to SQLAlchemy and after model optimization, cut San Francisco trees import time from an hour to 35 seconds.
Google introduced the Gemini 3.8 Flash and 3.8 Flash Cyber models. While the base Flash optimizes general agents, the Cyber version branches off as a specialized track for security operations. This shows that a one-size-fits-all model is beginning to hit its limits in the enterprise sector.
Google is rolling out WeatherNext 3, its latest global AI weather forecasting model. Instead of relying heavily on physics simulations, it feeds on live satellite data, narrowing predictions down to a 5-kilometer resolution updated hourly.
Anthropic is releasing Claude Fable 5.1 and its secure enterprise variant, Mythos 5.1. The major update includes a 75% price drop for cache reads to make long-running agents affordable, along with the model's ability to retain context over dozens of steps.
Sebastian Raschka demonstrates that you do not need big tech infrastructure to build a working AI detector. Using a custom dataset and a DistilBERT model, he built an API that reveals the limits of detection.
Alibaba released a new 27-billion parameter model under an Apache 2 license. It is the perfect size for running on a laptop, but due to its default setting, it suffers from massive verbosity on trivial tasks.
Google has added the ability to dynamically scan videos as an agent to its Gemini Flash models, without needing to send all frames at once. This improves the accuracy of image understanding, but mainly it radically reduces token consumption and overall costs.
OpenAI has overhauled Codex's architecture to maintain context and solve complex multi-step tasks. For development teams, this changes what is practically delegated and what still requires human approval.
Microsoft has distilled its massive pathology models into the Flash family. This delivers a 50x reduction in compute cost for whole-slide analysis, unlocking the possibility of repeated experiments across large patient populations.
Google's next-generation time-series model (TimesFM-3) introduces multivariate forecasting. Instead of slow step-by-step generation, it resolves the entire horizon at once, seamlessly incorporating future events like scheduled promotions or holidays.
The hype surrounding the massive deployment of the open-source agent OpenClaw in China was based on flawed data. Today, the project is largely ignored as development shifts toward managed cloud solutions.
