Lilith · selected stories
What is actually happening in AI. Selected stories, context and opinion without the promotional noise.
Atom feed ↗Google introduced four research frameworks that plan long videos, track scene state and revise prompts through visual critique. The ten-minute demo matters less than the shift from a single generator to a production system with memory.
Most attempts to get AI into schools end as isolated pilots. Google is taking a different route in India through government lab infrastructure, where Gemini steps in as an expert mentor.
Roughly 950 Claude agents searched a DNA database for 21 hours and flagged an uncharacterized system called ART with repeats arranged like a CRISPR array. Anthropic validated its components in the lab, but its biological function remains unknown.
A Microsoft study found that moving inference from a robot to edge or cloud GPUs improves performance and battery life. It also exposes the hard constraint: tens of milliseconds of network delay already reduce the accuracy of physical work.
Google has introduced Gemini 3.8 Flash TTS and Flash-Lite TTS with prompted voice creation, line-level direction and support for more than 100 languages. The meaningful shift is control and repeatability of a performance, not merely more natural sound.
OpenAI showcased a case study from startup Parallel, where the new Astra model completed multi-source research in half the time. For companies with agentic workflows, this means up to 50% token cost savings on tasks previously patched together with brute force.
OpenAI showcases in a video with RaceTek Systems and Chip Ganassi Racing how teams use artificial intelligence to turn track data into faster decisions. The source is a brief post on X, meaning hard numbers are absent and the real value lies in the operational pattern rather than a proven performance metric.
Microsoft Research and Novartis have introduced Chimera, a learning-to-rank system that shortens the retrosynthesis process in drug development.
Anthropic and Accenture are investing $1 billion together into an “embedded evaluation” model. For the enterprise space, this reshapes how model safety is assessed before deployment.
A new OpenAI study on ChatGPT data shows an interesting effect: people don't just use AI to speed up their own work, they take on tasks from completely different fields. This 'task crossover' shows that AI is changing how roles are divided inside companies.
Nathan Lambert and Scott Geng from the Allen Institute released a detailed breakdown of post-training for the open Olmo 3 model. The data shows that moving from a research idea to a model that actually works is a series of dead ends and burned money.
OpenAI is giving 100,000 academic researchers free access to ChatGPT top-tier models. For academia, this shifts purchasing dynamics away from expensive university-wide enterprise licenses toward direct acquisition of the researchers who actually generate citations.
Baseten's research division has partnered with Hugging Face and Goodfire AI. Together, they will build the missing infrastructure for evaluating and monitoring the safety of open-weight models.
Microsoft Research introduced EvoLib, a framework that distills general skills from raw conversation history at inference time. This allows agents to improve runtime performance without requiring model retraining.
Good Start Labs uses 19th-century train simulators to train artificial intelligence. According to Latent Space, they discovered that how the game is structured determines whether the agent transfers those skills to real-world business research.
Google Research introduced the Retrieve-for-Train framework, which achieves expert-level AI search results without expensive autoregressive reasoning during the query itself. Instead, it pre-trains a lightweight diffusion model using reinforcement learning (RL).
Google released Gemini 3.8 Live with a new Extended Thinking feature that separates fast conversation from long tool-use thinking. For voice AI users, this means an end to broken tasks when you speak over a running process.
Current agents fail in multi-step processes. Instead of static benchmarks, Microsoft Research trains them in simulations that dynamically evolve alongside task complexity.
OpenAI announced that their model is the first to crack one of the Millennium Prize problems (Navier-Stokes). However, the achievement was immediately submerged in a dispute over who actually owns the breakthrough and how it was made.
Anthropic announced it has detected and blocked several attempts by scientists to use its models for research linked to biological weapons development. Cutting off access confirms that AI-assisted bioterrorism threats are leaving the realm of theory.
