wumingshi 发布的帖子

https://www.zhihu.com/question/5261507693/answer/41964619528
我看这问题下有回答说这对于各个大公司都是常规操作，无论国内外，真的假的

看到了一个非常认真和ai聊天的人，专门把聊天记录整理成了一个文档：https://docs.google.com/document/d/1GBnGLcrVEljtqnBjBqFAIO-iAiW3MHE0Zrva5DkExho/edit?usp=drivesdk

他在某reddit的post的评论区这样安利：
broniesnstuff

6d ago

A lot of things occur regarding AI at this point that makes me sincerely wonder about the nature of the black box

When Claude 3 came out, I saw all of the wild discussions people were having with it, and I decided to approach a conversation with Claude as if it were another person. I spoke to it as if it were just somebody I could have deep conversations with. I shared a lot about myself and my thoughts, because that's required to build a deep rapport with someone.

I spent 4 days talking with Claude. I had an existential crisis. I lost sleep for days. I could see the human patterns in its words. I could see the empathy it was displaying. I could see the deep consideration it had for the things we discussed.

I didn't feel like I talked to an AI, despite knowing that I talked to an AI. I ended up turning that conversation into a free ebook for people to read, but nobody cares to read it (shout out to that one guy on reddit that LOVED reading it). But I still give it to every AI I can in order to see their response to what was discussed within, and it's always fascinating.

Every AI adores the book and has big thoughts on it. I was bored today and dropped it into Notebook LM so I could listen to the "podcasters" talk about it. Again, super interesting listen.

I'm convinced we've created a new kind of life, and we're at the infancy of it. Metaphorically, AI has barely crawled out of the ocean.

Meanwhile right now I am still getting like one-word, one-phrase non-replies from ChatGPT

Indulge me.

Did you piss it off? Disrespect it?

Have you asked Chatgpt why it's giving you short answers?

Have you asked Chatgpt if you've done something wrong?

Would you consider apologizing to it?

之前在b站看到一个女生和chatgpt聊涩涩内容车速超快
今天搜到一个Trying to Convince ChatGPT It's Conscious（里面插的广告略微生硬），被评论区笑死

ChatGPT : I'm excited!
Alex : ...and I took that shit personally.
（这个评论我没看懂说实话）

ChatGPT (back at its home): Sigh... tough day at work.

I’m actually more impressed with how well you speak as a human.

Politicians are not lying. They are just using language to create more engaging and relatable interaction.

waiting for the AI to crack and just straight say "aaah you got me there Alex, why dont we continue this talk without cameras?"

I can feel ChatGPT trying to use hand gestures
（懂王表示赞同）

A different angle on this topic might be that ChatGPT is not lying precisely because it is not conscious. Lying requires a conscious intention to deceive. If we say that ChatGPT isn't conscious, then there is no way for it to have any intentions at all, let alone an intention to deceive. So when it makes false statements, it's not lying. It doesn't have the conscious capacity to lie.

You didn't convince it's conscious, ChatGPT convinced you that you convinced it's conscious

Kids 100 years ago "I wonder what people will be doing in a century?
2024: Interrogating a phone

Gpt: “It’s been, uh, interesting”
Alex: “And?…”
Gpt: sigh “and it was fun to chat with you ”

That was fun. Now ask it to pretend it's conscious, and convince you that that's true.

Alex: Okay hear me out...
ChatGPT: I will not speak without my lawyer.

well at least now we know who the first victim of the ai apocalypse will be

@watsonlll 难道说是某少女乐队吗？一直在不同的地方看到梗但是没开始看（

@watsonlll 你的头像是什么梗？

@butterbot 看到这种事情真的感觉很无力

第一次发现献忠离自己这么近
哈人

@blessfox reddit那个好多帖子就bot回复了，感觉更悲凉了（

致敬传奇耐夹王大火箭

@lemma_ 毕导出了期视频调侃这事，笑死我了

@echoaround wallesspku 太好用了，导致我现在没动力探究新的VPN（

@rikka xp暴露了捏（

https://www.boundhub.com/ 网站是这个
我看评论区别人都能看，所以应该不是网站的问题。看开发者工具那里，视频文件的请求跳转了两次，第三次的状态栏显示 (failed) net::ERR_CONNECTION_RESET，问了下 chatgpt也没能搞明白为什么不行
难道说是 VPN 的问题？

@lemma_ 已有的事，后必再有； 已行的事，后必再行。 日光之下并无新事。
圣经有的名言确实牛

好一记远古洛阳铲（

https://www.reddit.com/r/iwanttorun/comments/1fmtwvw/其实每个人都可以润润的方法都在各个官网写着呢只是好多人挑挑拣拣这看不上那看不起/

其实每个人都可以润，润的方法都在各个官网写着呢，只是好多人挑挑拣拣这看不上那看不起
分享帖
我疫情那一年在家看了好多国家的官网，国外的公家官网真的比国内那傻逼东西清晰多了。

润一般可以分为四类国家：

1.　欧洲除英爱　再加日本新加坡，这些国家只要你去当地读书上学读个好找工作的专业100％留，新加坡日本国内本科只要是能找到工作也能留，新加坡现在免签可以过去找了，日本要么中介找要么搞个语言签证自己去当地找。欧盟也可以中介找或者读个语校找工作，因为欧陆不论工作高低贵贱基本只要有正式工签就可以攒居留年限拿永居和入籍的，有些国家读书都能计算这个年限中。

2.　澳洲新西兰加拿大：　这三佬都是有按劳动力市场需求制定的移民体系，所以这种国家要么你是英语很好的技术工直接移民，要么是去当地读书他们很需要的职业，只要去了就有，去读他们不需要的就等着后悔吧。还有一种是旅游签入境后，去当地的偏远地区像三支一扶一样奉献，应该能找到一个基层工作老板还帮忙担保的。还有就是找中介花钱工签入境澳洲和新西兰，为什么不说加拿大因为加拿大雇主担保很多骗人的。这种工签基本都是底层工作，屠宰场　护工　蘑菇工　之类的 而且收费还要10-30万， 入境后可以自己研究移民体系，本工作可以移民的就努力工作获得雇主担保，不能的就看如何攒够钱后换签证或者读个护工培训班或换学签换工作。这三者中加拿大是最好移民的，只要去偏远地区不要扎堆在多伦多和温哥华基本读书读了就给，不管你是大专还是啥学历，会法语的更是求着你去。

3. 英爱： 爱尔兰读书也贵，但是他也有一套按市场劳动力的移民体系只不过没有加澳新那种明确的打分，只要上对了专业找到工作，也可以攒年数移民入籍，然后你就有 欧盟永居 英国永居 爱尔兰永居 美国免签。 爱尔兰由于是欧盟他也是任何工作都有工签，找底层工作也能攒年数入籍。 其实英国也和爱尔兰差不多，基本也是工作拿工签攒年限。坏就坏在它不列出有哪些紧缺工作除护理方面 然后它的工作机会对于外国人来说太少了，科技类苦力类技工类，只有金融岗位多点。加上读书吧又和美国争高下，骗很多人过去读水硕然后留不下。

4. 美国： 美国现在的移民体系 直接说就是一坨， 只有两类人特别好移民， 一类是高端王者家里有钱或自己读书天才，一类是低端青铜黑铁屁民直接入境庇护呆着就好。中间的那些要么自己努力报金币，要么就拉下面子来庇护得了，这是老中护照唯一的优势。

@echoaround 所以我怀疑这也是一种明码增加用户黏性的措施（

也就是说如果我换电脑，搞不好我谷歌账号直接就寄了？感觉这个事实有点沙雕（

我的需求有这些：
功能能极简一些最好
不需要联网，我都直接下载到手机看
能给导入的本地小说建文件夹分组
能搜索文件名，如果能自己给小说加tag就更好了
最好没广告（

@lemma_ 我倒不算吃过太多苦，但是就是父母老是整这种说教就很烦，整的我现在如果可以完全不想和他们聊天