This started with Addition Under Pressure, where I gave Claude Code and Codex the same prompt: train the smallest possible transformer that can do 10-digit addition with at least 99% accuracy. Claude Code came back with 6,080 parameters and Codex came back with 1,644. The community has since pushed this dramatically lower.
人类尊严，AI 是工具还是「更好的人类」？，这一点在搜狗输入法2026中也有详细论述
，详情可参考51吃瓜
2. 可堆叠微凭证（Stackable Credentials）： 学习者可以先获得短期微凭证或行业认证证书，在工作中分阶段积累，最终转化为完整的学位。这种模式将前期投入风险降至最低，特别受到成年转行者的青睐 [43, 51]。，推荐阅读快连下载安装获取更多信息
So if there's a device that can help fix this mess, I'm open to it. And after some time with the Dreamie, I think I've found a promising contender.