德克萨斯扒鸡

@JiannrongY @wangzhian8848 简而言之，台湾和对岸现在还是处于战争状态，不是抢不抢的问题，未来一定会有大规模武装冲突爆发的一天，直到分出胜负。希望你未来自求多福。

@wangzhian8848 一個中國就是中華人民共和國，台灣是中華民國，這需要妥協什麼？我們沒有要搶中國代表權，為何你們一定要我們搶？

赖清德最新的讲话有没有妥协和改变？我认为是完全没有的。强调“对等尊严”这个赖清德的立场一直如此，但赖清德的对等的前提，是互不隶属，在他看来只要有一个中国的前提，就不存在对等和尊严。这次讲话里他也反复强调互不隶属，就是这个意思。但这种“对等尊严”在大陆看来其实就是台独。 从大陆视角看，什么叫妥协？当然最好是回归九二共识，这个对民进党比较难，再次就是回归中华民国宪法，强调中华民国是主权独立国家，不讲什么互不隶属，讲两岸不讲两国。达不到这两点，就都不算什么改变。 但对比赖清德上任以来的讲话，在语气上还是有些变化，现如今总是强调不能武力改变现状，其实就是看到大陆在加大武统准备，而川普眼看着又指望不上，口气明显不像刚上任那么任性了。 我记得当时赖清德眉飞色舞地讲，既然大陆想收复失地，去收复海参崴吗，干嘛对着我们台湾。这话现在他不敢讲了。他知道现在武统也就是大陆领导人的一念之间，也许哪句话没说好，习近平一生气，他一觉起来就被大陆海警远洋捕捞，跨省逮捕了。

@MaaSonder The lockout period during this time will definitely need to be extended. Otherwise, Anthropic is really going to ruin its reputation with the general public.

Even with Fable 5 returning, is the subscription cutoff still going to be locked to June 22nd?

@igeekbb 东施效颦

逐帧学习！

@LaurenceMister 雷军现在做的就是将个人形象逐步和小米脱钩，避免个人负面新闻波及整个公司和品牌形象。做终端服务和产品的公司很多，没见哪家老板这么出来抛头露面

雷军在武汉早上吃个面，就能让全体小米股东也跟着一起吃面。

@ChingteLai 日杂，怎么想起来过中国的节日？

Sending warm #DragonBoatFestival greetings to everyone celebrating in #Taiwan & around the world! A big thank you to all those on duty during the holiday. As you unwrap your favorite zongzi today, may you also unwrap endless joy, good health and success ahead!

@LaurenceMister 雷军这两年愈发的像岳不群，公司产品做的不伦不类公司股价萎靡不振，本人也虚与委蛇道貌岸然装模作样。

我有点生理不适了，这也要同步吗？

@rwayne 但凡用过，都知道这个软件有多垃圾

我说一个暴论，如果搞AI你要商业化 今年下半年，必须去研究workbuddy 不管你是卖课还是做交付 没有为什么，也别问为什么！

@Jackywine 可以实现无限上下文，但你甭管对不对就是了

小米 MiMo Code 这个无限上下文咋实现的啊？？

@thsottiaux Alright, it's time for a nationwide promotion to reset credit limits

Oy. We are aware that some Codex users are experiencing high error rates with "model at capacity" and are working to bring things back to being stable. status.openai.com

@cuichenghao 仓廪实而知礼节，衣食足而知荣辱。

有网友在工地附近直播卖盒饭，有粉丝送给他一箱红牛，他决定送给买盒饭的工人，但前提是对镜头说一声谢谢，感谢一下馈赠人就行。有一个工人死活不肯说谢谢，不让拿就摔了盒饭、打翻饮料还动手打人。该网友问，以后还敢做好人吗？

@Tristanrhee3 The new and distinct advantage lies in how we leverage AI tools to better optimize workflow efficiency and maximize output.

Everyone has AI now. That advantage expired fast. What's the new advantage?

@francoisfleuret India's caste system is deeply ingrained and cannot be changed after birth. In contrast, China's household registration system (hukou) is something that can be changed later in life. These are two completely different things.

Removed a re-post of a tweet about castes in China, since the only caste-like discrimination that seems to remain formally in China is based on locality registration, and boils down to rural vs. urban origin which is bad, but far from castes as in e.g. India.

@rezoundous If AIGC eventually reaches a level of intelligence that far surpasses the collective sum of human intellect, we will likely see the emergence of various religious sects centered around it. This process would be quite similar to the way the Trisolaran religion

Anyone feeling AI getting cult-ish?

@rezoundous Aside from the fact that it burns through compute and tokens a bit quickly, this is definitely the most practical model available right now. Of course, the high consumption rate isn't the model's fault: it's just because we're too poor.

Fable 5 is the first model I've seen that has near unanimous positive feedback on X

@Xudong07452910 小米目前还没资格喷这几家，模型垃圾一米

小米工程师嘲讽某国产大模型，好难猜是哪一家啊？😂

@jianghaikuo666 当你在思考下载之后需要干什么的时候，其实你就不用用这个东西了

我是真有点纳闷，你们普通人平时到底都拿 Codex 干什么？ 我充 Plus 都快两个月了，Codex 只认真用过一次，做了个桌宠，之后再打开就完全不知道该让它干什么。 有问题直接问 GPT 基本就够了，我又不是程序员，也没有一堆机械化、重复性的工作需要它帮我自动处理。 每次看到别人说 Codex 多强、多方便，我都觉得它确实厉害，但这种厉害好像跟我没什么关系。 第二个月 Plus 都快到期了，我甚至还没找到第二个真正需要用 Codex 的场景。 所以我很好奇，你们这么多人都说用codex，但不说用来做什么，不写代码的普通人，到底都是怎么用 Codex 的？还是你们就单纯写个帖子发出来蹭个流量？

@O_Anu_O In fact, except for a very few areas, China has actually followed a very standard path of fully competitive market economy

Ironic how the so-called "Communist" country China has all the open source AI models, and a democratic country US forces shut down of AI models.

@thsottiaux I've successfully recommended two friends to sign up for Codex

Hi, I'm Tibo and I just discovered Codex. AMA

@Angaisb_ You're overdoing it with the show-off

This is what happens when you keep telling everyone how "dangerous" your model is for months and then suddenly release it because it never really was that dangerous and you just needed more compute Anthropic kinda did this to themselves

@saltyAom I laughed my ass off

If a government can dictate your company what or what not to do then how is it any difference from CCP at all