Measuring AI Ability to Complete Long Tasks
(news.ycombinator.com)
1801.
1802.
Measuring AI Ability to Complete Long Tasks: Opus 4.5 has 50% horizon of 4h49M
(news.ycombinator.com)
1803.
1804.
The dawn of a world simulator
(news.ycombinator.com)
1805.
1806.
1807.
1808.
1809.
1810.
The LCD Steam Deck is done
(theverge.com)
1811.
What Does a Database for SSDs Look Like?
(news.ycombinator.com)
1812.
Reflections on AI at the End of 2025
(news.ycombinator.com)
1813.
Building venture-backable companies in heavily regulated spaces
(techcrunch.com)
1814.
1816.
1817.
1818.
1820.
1821.
1822.
1823.
1824.
1825.
1826.
History LLMs: Models trained exclusively on pre-1913 texts
(news.ycombinator.com)
1827.
Trained LLMs exclusively on pre-1913 texts
(news.ycombinator.com)
1828.
1829.
1830.
Today's top topics:
google
apple
openai
google health
anthropic
samsung
android authority
chatgpt
nvidia
spacex