←搜索结果

搜索“Microsoft”

按匹配度 · 共 1144 条结果 清除搜索

星期日

  1. Benedict Evans：Blog

    AI 伦理的问题——Benedict Evans

    Benedict Evans 以英国邮局富士通系统冤案为例指出，AI 伦理与 AI 监管面临的核心难题是抽象层级错误：不同场景的 AI 应用（假释、房贷、内容审核、药物研发）是完全不同的问题，不存在一个能覆盖所有情况的伦理领域或统一监管法律。他同时提到 Google Gemini 争议，并强调生成式 AI 的形态连业内都难以预测，立法需要保持谦逊。

    a SQL regulator, and the same point applies when we read that FTX had a spreadsheet with eight different balance sheets. We don’t call for a spreadsheet regulator, and demand that Microsoft stop this from happening again. That would be the wrong level of abstraction. This, I think, is the challenging in talking about ‘AI Ethics’, or in writing laws to regulate ‘AI’. There are lots of ways that bad people can do bad things with software, and there are lots of ways that people can screw up with so

    匹配字段正文

    #analysis#行业动态#政策/监管

星期二

  1. a16z：News

    a16z 创始人 Marc Andreessen 长文：AI 将拯救世界

    a16z 联合创始人 Marc Andreessen 发表长文《Why AI Will Save the World》，主张 AI 不会毁灭人类、不会摧毁社会、不会造成大规模失业，反而会带来生产力与经济繁荣。他把 AI 风险恐慌类比为历史上的技术道德恐慌，用 Baptists 与 Bootleggers 框架批评监管推动者，并称唯一真实风险是中国取得全球 AI 主导地位，主张美国全力加速 AI 发展、放开开源与创业竞争。

    an afford them. This is precisely what is already happening in AI – it’s why you can use state of the art generative AI not just at low cost but even for free today in the form of Microsoft Bing and Google Bard – and it is what will continue to happen. Not because such vendors are foolish or generous but precisely because they are greedy – they want to maximize the size of their market, which maximizes their profits. So what happens is the opposite of technology driving centralization of wealth

    匹配字段正文

    #analysis#行业动态#OpenAI#大佬观点

星期四

  1. a16z：News

    a16z 发布 AI Canon：现代 AI 必读资源精选清单

    a16z 发布“AI Canon”精选资源清单，汇集过去几年对 AI 领域影响最大的论文、博客、课程与指南。清单从 Transformer 与潜在扩散模型入门讲起，延伸到 LLM 技术学习、实践指南与 AI 市场分析，并收录以 2017 年 Google 论文《Attention is All You Need》为起点的里程碑研究。

    ng development timeline, size, training strategies, training data, hardware, and more; Sparks of artificial general intelligence: Early experiments with GPT-4: Early analysis from Microsoft Research on the capabilities of GPT-4, the current most advanced LLM, relative to human intelligence; The AI revolution: How Auto-GPT unleashes a new era of automation and creativity: An introduction to Auto-GPT and AI agents in general. This technology is very early but important to understand—it uses intern

    匹配字段正文

    #analysis#其他

星期四

  1. Lilian Weng：Lil'Log

    Lilian Weng 综述广义语言模型：从 CoVe、ELMo 到 GPT-2、BERT

    Lilian Weng 在 Lil'Log 发布《Generalized Language Models》综述，梳理了让词嵌入依赖上下文并更易迁移到下游任务的各类方法。文章从 CoVe 的 NMT 编码器上下文词向量讲起，重点介绍 ELMo 通过无监督双向语言模型（biLM）预训练学习上下文表示，并指出 OpenAI GPT 与 BERT 进一步用无监督预训练加生成式架构同时克服 CoVe 的局限。该文后续陆续补充了 ULMFiT、GPT-2、ALBERT、RoBERTa、T5、GPT-3、XLNet、BART 和 ELECTRA。

    argument structure of a sentence, and is often described as answering “Who did what to whom”. CoNLL-2004 & CoNLL-2005 Sentence similarity: also known as paraphrase detection MRPC (MicRosoft Paraphrase Corpus): It contains pairs of sentences extracted from news sources on the web, with annotations indicating whether each pair is semantically equivalent; QQP (Quora Question Pairs) STS Benchmark: Semantic Textual Similarity Sentence Acceptability: a task to annotate sentences for grammatical accept

    匹配字段正文

    #analysis#教程/实践#数据/训练#OpenAI