92. Show HN: Stickerbox, a kid-safe, AI-powered voice to sticker printer
93. A tagging system for documentation review comments
94. Response Healing: Reduce JSON defects by 80%+
95. How getting richer made teenagers less free
96. Open Source Historical LLM trained exclusively on 19th century text
97. GPT-5.2-Codex
98. Chinese molten salt reactor achieves conversion of thorium-uranium fuel
99. Meta Segment Anything Model Audio
100. Noclip.website – A digital museum of video game levels
101. Designing a Passive Lidar Detector Device
102. Please just try HTMX
103. Sequoia's Shaun Maguire accused an innocent Palestinian of being Brown shooter
104. Two kinds of vibe coding
105. I've been writing ring buffers wrong all these years (2016)
106. Getting bitten by Intel's poor naming schemes
107. Nvidia Is the Only AI Model Maker That Can Afford to Give It Away
108. The Scottish Highlands, the Appalachians, Atlas are the same mountain range
109. T5Gemma 2: The next generation of encoder-decoder models
110. How China built its ‘Manhattan Project’ to rival the West in AI chips
111. We ran Anthropic’s interviews through structured LLM analysis
112. Building a Transparent Keyserver
113. Classical statues were not painted horribly
114. FunctionGemma 270M Model
115. Show HN: Misata – synthetic data engine using LLM and Vectorized NumPy
116. Valve is running Apple's playbook in reverse
117. Available to download Friday, some Epstein files no longer there Saturday
118. Working quickly is more important than it seems (2015)
119. How to hack Discord, Vercel and more with one easy trick
120. Debian adds LoongArch as officially supported architecture
- Previous Page Next Page