一个名为 Talkie 的新项目发布了一个仅使用 1931 年前英文文本训练的 130 亿参数语言模型。这个“复古”模型旨在探索人工智能预测未来和生成超出其训练数据截止日期的新颖想法的能力。虽然基础模型是开源的，但经过微调的聊天版本依赖于 Claude Sonnet 和 Opus 等现代大型语言模型进行辅助，这引发了对潜在时代错乱污染的担忧。 AI
影响 为研究人工智能泛化和历史知识表示提供了一个独特的研究工具。
排序理由 发布了一个基于历史数据训练的新语言模型，具有研究意义。
AI 生成摘要 · Google Gemini · 来自 9 个来源。 我们如何撰写摘要 →
<p><strong><a href="https://talkie-lm.com/introducing-talkie">Introducing talkie: a 13B vintage language model from 1930</a></strong></p> New project from <a href="https://nlevine.org">Nick Levine</a>, <a href="http://www.cs.toronto.edu/~duvenaud/">David Duvenaud</a>, and <a href…
<p><a href="https://lobste.rs/s/uws0nc/introducing_talkie_13b_vintage_language">Comments</a></p>
Introducing talkie: a 13B vintage language model from 1930 https:// talkie-lm.com/introducing-talk ie # AI # Talkie
Introducing talkie: a 13B vintage language model from 1930 https:// lobste.rs/s/uws0nc # ai https:// talkie-lm.com/introducing-talk ie
Introducing talkie: a 13B vintage language model from 1930 "Have you ever daydreamed about talking to someone from the past? What would you ask someone with no knowledge of the modern world? What would they ask you? While we don’t have time machines yet, we can simulate this expe…
Thoughts on Historical Language Models and Talkie-1930 https://resobscura.substack.com/p/are-vintage-llms-the-start-of-a-new # HackerNews # Tech # AI
Introducing talkie: a 13B vintage language model from 1930 https://simonwillison.net/2026/Apr/28/talkie/#atom-everything # AI # VintageTech # LanguageModel
Talkie: a 13B vintage language model from 1930 https:// talkie-lm.com/introducing-talk ie # HackerNews # Talkie # 1930s # Vintage # Language # Model # AI # Technology # NLP
Talkie: a 13B vintage language model from 1930 https://talkie-lm.com/introducing-talkie # HackerNews # Tech # AI