On some level, calories in calories out has to be true. But these variables are not independent. Bodies respond to exercise by getting hungry and to calorie deficit by getting tired. Even absent th…

Now in early beta for SuperGrok Heavy subscribers — Grok Build is a new coding agent that runs right from your terminal.

Home page of Josh Horowitz, UW PhD researcher

A significant proportion of queries to large language models ask them to edit user-provided text, rather than generate new text from scratch. While previous work focuses on detecting fully AI-generated text, we demonstrate that AI-edited text is distinguishable from human-written and AI-generated text. First, we propose using lightweight similarity metrics to quantify the magnitude of AI editing present in a text given the original human-written text and validate these metrics with human annotators. Using these similarity metrics as intermediate supervision, we then train EditLens, a regression model that predicts the amount of AI editing present within a text. Our model achieves state-of-the-art performance on both binary (F1=94.7%) and ternary (F1=90.4%) classification tasks in distinguishing human, AI, and mixed writing. Not only do we show that AI-edited text can be detected, but also that the degree of change made by AI to human writing can be detected, which has implications for authorship attribution, education, and policy. Finally, as a case study, we use our model to analyze the effects of AI-edits applied by Grammarly, a popular writing assistance tool. To encourage further research, we commit to publicly releasing our models and dataset.

I built an app that makes language models paint iteratively to explore whether the process produces something more artistically sincere than one-shot generation.

In which it takes me three years to get round to testing my code

This PR has been marked as AI slop and the description has been updated to avoid confusion or misleading reviewers.
Many AI PRs are fine, but sometimes they submit a PR too early, fail to test if t...

Web architecture is built on a 20-year-old assumption that state lives in the database, and compute is stateless. But we

Instructure, Canvas’ parent company, said it “reached an agreement with the unauthorized actor involved in this incident.”

We're changing the name of our company to Fin. Intercom will live on as our customer service software platform, but Fin is our future.

Deathbed speeches by Henry James about the Napoleonic legend

Wrap Go binaries in Python wheels. Contribute to simonw/go-to-wheel development by creating an account on GitHub.

With a style that was both accurate to the work and approachable to the reader, Ned Rozell authored the Alaska Science Forum for 31 years.

I was on the bus on my way to an important business meeting about the exponential growth of my MRR, when suddenly I noticed some troublemakers were hassling an old lady. Their caps were on backwards, their clothes were baggy, and probably thought the grindset was about skateboarding.
I had to act. I opened my macbook and showed Claude what was happening. “Claude, we gotta do something!”, I said. “You’re absolutely right!”, Claude affirmed in a calm and confident font, and thanks to --dangerously-skip-permissions, it had started coding before I had even thought about what to do. It was “cannoodling”, which I think is one of the good ones.

Runo extracts structured, typed JSON from any URL using AI. Define your schema, call the API, get clean data. Handles JS rendering, bot bypass, and batch crawling.