Tag
OCR researcher Vik Paruchuri highlights the dramatic progress in handwriting recognition models over the past two years, noting that outputs have gone from barely readable to nearly perfect transcription.
The article reflects on the rapid progress in AI models from three years ago, comparing early models like Bard that struggled with coding tests to current models like Qwen 27B and Opus 5.5, and speculates on future advancements.
A researcher reports significant but incomplete progress on the Hodge conjecture, a famous unsolved problem in algebraic geometry.
The article discusses an achievement in technology or AI that was completed in just two years.
François Chollet, once an AI skeptic, has revised his AGI timeline to expect it sooner than ~2030 due to faster-than-anticipated progress in the field.
The article hints at accelerating progress in technology or AI, suggesting that advancements are moving faster.
The tweet expresses enthusiasm for AI research opportunities, particularly with AI agents, and highlights the rapid progress in machine learning.
Discusses Grok 4.6's position on a chart and asks for opinions on Grok's progress.
Elon Musk reflects on how much has changed in the past 10 years and asks followers to imagine what the next decade will bring.
Elon Musk comments on the latest heat shield design for SpaceX Starship, highlighting improvements across flights 10-13 and noting that progress is being made.
George Lucas compares rejecting AI to rejecting cars in favor of horses, arguing AI is the future and inevitable. He dismisses concerns about AI, equating resistance to technological progress to ignoring past advancements.
The article argues that computer-use/browser-use AI capabilities are progressing very quickly and will agentify the web, as most of the web lacks APIs.
This paper presents a comprehensive overview of metacognition in large language models, covering measurement methods, improvement techniques, and future directions.
Observing that new AI models are being shipped every two weeks, but the bottleneck has shifted elsewhere, implying that the challenge is no longer just model development.
A reminder of how far AI video generation has advanced in just three years, highlighting the rapid progress in the field.
The article examines the current state of quantum computing, noting that no quantum computer has yet performed a useful task despite ambitious promises from the Trump administration and Microsoft's Majorana 2 chip announcement, which has drawn criticism from independent researchers for overhyping incremental progress.
An observation that future technologies often start as seemingly trivial toys before evolving into foundational infrastructure for civilization.
A reflective piece asking what recent AI developments would have seemed most unbelievable in 2020, and what future surprises might await.
This paper revisits the WorkBench benchmark for workplace agents two years after its initial release, showing that the best agent (Claude Opus 4.8) now completes 89% of tasks with only 2.5% harmful side effects, compared to GPT-4's 43% completion and 26% harm rate in 2024. It finds that capability and safety improve together, open-weight models have drastically lowered costs, and some basic mistakes persist.
Discussion of recent AI model scores on the 'humanity's last exam' benchmark, noting improvement from GPT-4o's 2.7% in May 2024 to around 45% by June 2026, questioning the exam's difficulty.