| 留园首页 新闻速递首页 热评新闻榜 新闻评论公告 72小时神评妙论 即刻热度新闻 |
[2楼]评论人:XXyourOO
噗，黄皮犬粮真是抓住一个口误能吃好几年的福利屎 🤣
噗，黄皮犬粮真是抓住一个口误能吃好几年的福利屎 🤣
傻逼不会问问题。claude.ai 很简单给出正确答案
回复22楼:If there was a report or instance where ChatGPT provided an incorrect comparison, such as stating that 9.11 is bigger than 9.9, it would likely be due to a misinterpretation or an error in processing the input. Properly, 9.11 should be compared to 9.9 as follows:
- 9.11 is smaller than 9.9 because when comparing decimal numbers, you look at each digit from left to right. Here, 9.1 (from 9.11) is clearly less than 9.9.
Errors like this can occur in complex AI systems, and it's important for developers to continually refine the algorithms and data interpretations to minimize such mistakes. Continuous feedback and updates help improve the accuracy and reliability of AI responses.
- 9.11 is smaller than 9.9 because when comparing decimal numbers, you look at each digit from left to right. Here, 9.1 (from 9.11) is clearly less than 9.9.
Errors like this can occur in complex AI systems, and it's important for developers to continually refine the algorithms and data interpretations to minimize such mistakes. Continuous feedback and updates help improve the accuracy and reliability of AI responses.
假消息….真会编….人傻还真试了.人不尴尬，尴尬的就是别人。有chat都可以去了解一下
研究证明，人脑的逻辑系统和语言系统是两个不同的大脑区域，人类没有语言的时候也能进行逻辑思考如孩童，有的人能流利交流却完全失去逻辑思考能力。动物不会语言也有行动逻辑。所以我认为大语言模型就是先天逻辑缺陷，再多的训练数据它也不会有真逻辑推理能力，更加没有的自我意识。
必须问：9.11 和 9.90 哪一个大？ 答案是9.90 大
语言和数学最难的是一词多用，ambiguity. The dot . 是可以表达为文书中的dot. 所以9.11 在表达为文章 section number, 的确是9.11 大于9.9。
但dot 表达为小数点 point 的时候，意义就不一样。 大模型在 language parser and token extraction 的时候是没有用neuro network. 是rule based 的， 所以谁能用LLM 搞个 language parser 就能解决这问题
但dot 表达为小数点 point 的时候，意义就不一样。 大模型在 language parser and token extraction 的时候是没有用neuro network. 是rule based 的， 所以谁能用LLM 搞个 language parser 就能解决这问题
这说明什么？说明AI训练还不够，需要更多更快地芯片，英伟达继续牛逼！
再次证明不懂逻辑只考死记硬背也可以一本正经说瞎话甚至不少难题也能蒙对。但不懂逻辑总归下盘不稳，它再厉害你也不知道它在啥时候会突然彻底弱智。不能委以重任。
做程序的人第一眼肯定说9.11比9.9大。习惯思维。AI现阶段还不能用于登月计算。
OpenAI经常不能正确地回答问题，或者解决问题。有时候指出它明显的错误，前后矛盾的地方，它只会一个劲地道歉，然后把错误内容再复述一遍，看不到一点智能。
所以tm现在有缺陷的人训练出来的AI果真同样的问题，高智商低常识，像不像疯子科学家？
牛逼啊，试了下，确实问题很大，让他计算一下0.11和9.9，竟然给出这种答案，真没想到，难怪今天英伟达大跌。
9.11 - 9.9 = 0.21
0.21 + 9.9 = 10.11
9.11 - 9.9 = 0.21
0.21 + 9.9 = 10.11
当然是9:11大！炸毁美国标志性摩天搂，炸死6000多人。应该够大
别的不知道，chatgpt的数学一直不好。经常简单的加减乘除都会算错。毕竟是语言模型，没加进去计算器。
用Perplexity试了一下，回答正确啊。
9.11 and 9.9——which is bigger?
Answer
To determine which number is bigger between 9.11 and 9.9, let's compare them:
9.11 = 9 + 0.11
9.9 = 9 + 0.9
The whole number part (9) is the same for both, so we need to compare the decimal parts:
0.11 < 0.9
Therefore:
9.11 < 9.9
9.9 is the bigger number.
To visualize this on a number line, 9.9 would be closer to 10, while 9.11 would be just slightly above 9.1.
9.11 and 9.9——which is bigger?
Answer
To determine which number is bigger between 9.11 and 9.9, let's compare them:
9.11 = 9 + 0.11
9.9 = 9 + 0.9
The whole number part (9) is the same for both, so we need to compare the decimal parts:
0.11 < 0.9
Therefore:
9.11 < 9.9
9.9 is the bigger number.
To visualize this on a number line, 9.9 would be closer to 10, while 9.11 would be just slightly above 9.1.
问题换成其他数字比较，比如8.11和8.9时，就不会出现这种错误。
顺便说一句，2L是个傻逼。
顺便说一句，2L是个傻逼。
我觉得文章说的有缺陷，在答案后的解释里，ai已经明确分析出对比的双方都是数字，却无法识别这个符号的实际数值。后面说道提示ai符号内容是double，ai就能识别其正确数值了。不提示ai就认为这是一种现实中不存在奇怪编码的数字。这说明在符号分析的过程中可能存在固定代码。
9.11像徵中東絕望中的反抗精神，9.9像徵資本家割韮菜方式。你說誰大？人會認為是資本家，中立的AI不會。
擦，9.11 > 9.9 这么明显的bug！这就是AI版的“通商宽衣”呀！😂
|
|
|
|
|
|
Copyright (C) 6park.com All rights Reserved.