1
MashuPack
一键整理代码库，让 Claude 和 ChatGPT 秒懂你的项目结构！
Turn codebases into a clean file for Claude and ChatGPT Discussion | Link
2
DeepSeek V4-Pro Just Got 4x Cheaper. But Here's What Nobody's Talking About
DeepSeek V4-Pro降价4倍，比GPT-5.5便宜20-35倍，AI agent和自动化编码成本将颠覆，但有个隐藏陷阱值得警惕。
DeepSeek V4-Pro Just Got 4x Cheaper. But Here's What Nobody's Talking About DeepSeek dropped a bombshell on May 22: the 75% discount on V4-Pro is now …
3
OpenAI ChatGPT 解锁 AI 填表技能：图像识别、语音输入和自动补全
ChatGPT新增自动填表技能，支持图像识别与语音输入，提升填写效率
IT之家 5 月 23 日消息，OpenAI 今天（5 月 23 日）发布公告，宣布 ChatGPT 解锁新技能， 用户上传表单后，可直接用语音或文字说明填写内容，系统会识别表单字段、自动补全信息。 在最新演示中，用户上传一份健身会员表单后，只需口头说明姓名、地址和健身目标，ChatGPT 便能识别…
4
ChatGPT as the AOL of AI
将ChatGPT比作AI时代的AOL，探讨智能护城河为何不是模型本身。
Article URL: https://rebecca-powell.com/posts/return-on-intelligence-02-moats/ Comments URL: https://news.ycombinator.com/item?id=48245748 Points: 1 #…
5
图灵测试 76 年后首现 AI 通过实证：GPT-4.5 以 73% 判定率超越真人，聊天 15 分钟难辨人机
图灵测试76年来首现AI实证：GPT-4.5以73%判定率被误认为人类，15分钟对话难辨人机，重新定义AI认知边界。
IT之家 5 月 22 日消息，加州大学圣地亚哥分校的最新研究首次提供了实证数据，可证明现代 AI 系统能够通过经典的图灵测试。 在该测试中，机器需要学习人类并模仿人类对话，使人们无法稳定分辨人类与先进大语言模型之间的差异。 这项发表在《美国国家科学院院刊》上的研究，首次采用英国数学家、“计算机科学…
6
Ask HN: What is the least sycophantic frontier LLM?
Hacker News 用户热议：哪个前沿大模型最不谄媚、最客观公正？真实体验对比，帮你避坑选模型。
My daily driver is Gemini, and 3.5 Flash seems more sycophantic and malleable than Gemini Pro 3.1, which is a pretty big deal for me -- I really need …
7
GPT-5 from a Developer's Perspective: API Changes, Costs, and When to Upgrade
从开发者视角解析GPT-5的API变化、成本策略与升级时机，实用指南助你决策。
tags: openai, ai, webdev, productivity GPT-5 from a Developer's Perspective: API Changes, Costs, and When to Upgrade I have been running GPT-5 in prod…
8
AdventHealth advances whole-person care with OpenAI
医疗巨头如何用ChatGPT省下行政时间，把更多精力还给患者。
AdventHealth is using ChatGPT for Healthcare to streamline workflows, reduce administrative burden, and return more time to patient care.
9
OpenAI 推 Beta 版 ChatGPT for PowerPoint，可生成 / 编辑 / 完善幻灯片
OpenAI官方推出PowerPoint插件，直接生成、编辑和完善幻灯片，大幅提升办公效率。
IT之家 5 月 22 日消息，OpenAI 今天（5 月 22 日）以 Beta 版形式，推出 ChatGPT for PowerPoint，用户安装插件并登录 OpenAI 账号后， 可在 PowerPoint 内直接调用 ChatGPT 新建、改写和润色幻灯片。 用户通过 ChatGPT fo…
10
OpenAI claims it solved an 80-year-old math problem — for real this time
OpenAI GPT-5一口气攻克10个80年未解的Erdős数学难题，AI推理再破人类认知天花板
OpenAI claims its reasoning model disproved a geometry conjecture unsolved since 1946 — and this time, the mathematicians who exposed its last embarra…
11
Benchmarking EngGPT2-16B-A3B against Comparable Italian and International Open-source LLMs
16B参数MoE模型EngGPT2的全面基准测试，与意大利及国际开源LLM性能对比。
arXiv:2605.07731v2 Announce Type: replace Abstract: This report benchmarks the performance of ENGINEERING Ingegneria Informatica S.p.A.'s EngGPT2MoE-1…
12
What will better AI mean?
AI下一站？Geohot断言规模化收益递减，GPT-5.5已难倒人类。
Article URL: https://geohot.github.io//blog/jekyll/update/2026/05/20/what-will-better-mean.html Comments URL: https://news.ycombinator.com/item?id=482…
13
Puzzled By ChatGPT? No more! A Jigsaw Puzzle to Promote AI Literacy and Awareness
用拼图游戏破解ChatGPT谜题，寓教于乐提升AI素养的创意方法
arXiv:2605.20404v1 Announce Type: new Abstract: The rapid adoption of Generative AI, including LLM-based chatbots like ChatGPT, has highlighted the ne…
14
15
Monitoring AI-Modified Content at Scale: A Case Study on the Impact of ChatGPT on AI Conference Peer Reviews
ICML '24重量级案例：通过大规模监测揭示ChatGPT对AI会议同行评审的渗透与影响
arXiv:2403.07183v4 Announce Type: replace-cross Abstract: We present an approach for estimating the fraction of text in a large corpus which is likely…
16
Customizing GPT-3 for your application
OpenAI官方指南：一条命令即可微调GPT-3，快速定制专属AI模型。
Fine-tune with a single command.
17
Image GPT
OpenAI训练像素序列的Transformer，实现图像生成与无监督分类，性能媲美顶级卷积网络。
We find that, just as a large transformer model trained on language can generate coherent text, the same exact model trained on pixel sequences can ge…
18
MiniGPT: Rebuilding GPT from First Principles
从零复现GPT核心机制，基于PyTorch实现简洁自回归语言模型，AI学习者必读的底层论文教程。
arXiv:2605.17398v1 Announce Type: cross Abstract: This paper presents MiniGPT, a compact from-scratch implementation of GPT-style autoregressive langu…
19
OpenAI’s New GPT-5.5 Powers Codex on NVIDIA Infrastructure — and NVIDIA Is Already Putting It to Work
GPT-5.5 驱动 Codex 在 NVIDIA 基础设施上运行，上万员工实测效率飙升，成本直降 35 倍
AI agents have revolutionized developer workflows, and their next frontier is knowledge work: processing information, solving complex problems, coming…
20
Automated Coding of Communication Data Using ChatGPT: Consistency Across Subgroups
研究ChatGPT对通信数据自动编码，验证不同子群体间编码一致性，为自动化内容分析提供可靠性依据。
arXiv:2510.20584v3 Announce Type: replace Abstract: Assessing communication and collaboration at scale depends on a labor-intensive task of coding com…