Tootfinder

“TABS [by Mozilla] pulls exactly the data you need—from HTML to Markdown to JSON—using the fastest, most efficient method for each page. It adapts to the structure and complexity of the site, staying stealthy and reliable so your [AI] agents always get what they need without friction.”
Ethical Stealthy AI Scraping (tm) by Mozilla.
#Mozilla

Today, I'm at Bundesarchiv in Koblenz for the Strategy & Planning meeting of our project "Wiedergutmachung". Our task in this project is to develop efficient information extraction from historical case files of the German Postwar recompensation process of nationalsocialist injustice.
@… @…

This evening I have been listening to one of @… 's podcasts and thinking about my failure in trying to lead the village's planning working group, and about the cognitive dissonance underlying my Tricycle project. I suspect this essay will be a grim read; it's not well formed in my mind as I sit down to write.

Google plans to invest $10B to build a new 1GW data center cluster near Visakhapatnam in Andhra Pradesh, India, with operations expected to begin by July 2028 (Shashank Pathak/Entrackr)
https://entrackr.com/snippets/google-to-invest-10-bn-in…

Google to invest $10 Bn in Visakhapatnam data centre cluster
Google is planning a massive push in India by investing $10 billion to build a new data centre cluster in and around Visakhapatnam. The project will include three campuses. Snippets

Back when I worked at SDC in Santa Monica (roughly 1971 through 1980) one of our department's projects was an early AI project.
We had an entire IBM 370 mainframe running CP/67 (which became IBM's VM) running LISP based AI code.
I think that the group may have been using it for continuous speech recognition - which we knew our "national security' customer was planning to use to automate wiretaps.
(I used the speech recognition project's soundproof room …

SITCOM: Scaling Inference-Time COMpute for VLAs
Ayudh Saxena, Harsh Shah, Sandeep Routray, Rishi Rajesh Shah, Esha Pahwa
https://arxiv.org/abs/2510.04041 https://

SITCOM: Scaling Inference-Time COMpute for VLAs
Learning robust robotic control policies remains a major challenge due to the high cost of collecting labeled data, limited generalization to unseen environments, and difficulties in planning over long horizons. While Vision-Language-Action (VLA) models offer a promising solution by grounding natural language instructions into single-step control commands, they often lack mechanisms for lookahead and struggle with compounding errors in dynamic tasks. In this project, we introduce Scaling Infere…

Score-based generative emulation of impact-relevant Earth system model outputs
Shahine Bouabid, Andre Nogueira Souza, Raffaele Ferrari
https://arxiv.org/abs/2510.04358 https://

Score-based generative emulation of impact-relevant Earth system model outputs
Policy targets evolve faster than the Couple Model Intercomparison Project cycles, complicating adaptation and mitigation planning that must often contend with outdated projections. Climate model output emulators address this gap by offering inexpensive surrogates that can rapidly explore alternative futures while staying close to Earth System Model (ESM) behavior. We focus on emulators designed to provide inputs to impact models. Using monthly ESM fields of near-surface temperature, precipitat…