ChatGPT火爆出圈，AI聊天到底有多像人类？
2022.12.23
肖云照
难度：CSE7
课程导读

最近，一款AI聊天机器人ChatGPT全网刷屏，它是由人工智能公司OpenAI推出的，而且发布不满一周，用户数就超过百万，又掀起了一波人工智能讨论的热潮。ChatGPT能写笑话、写代码、写论文，而且是用多种不同语言对答如流。这款新型聊天机器人的实用价值到底有多大？它的出现是否意味着AI发展进入了新阶段？一起来听今天的讲解。

本文封面图片来自techcrunch.com，如有版权问题，请联系删除。

本篇课程首发于2022年12月23日
打水漂矣

英文原文

The Brilliance and Weirdness of ChatGPT

By Kevin Roose

ChatGPT is, quite simply, the best artificial intelligence chatbot ever released to the general public. It was built by OpenAI, the San Francisco A.I. company that is also responsible for tools like GPT-3 and DALL-E 2, the breakthrough image generator that came out this year.

For most of the past decade, A.I. chatbots have been terrible — impressive only if you cherry-pick the bot’s best responses and throw out the rest.

But ChatGPT feels different. Smarter. Weirder. More flexible. It can write jokes, working computer code and college-level essays. It can also guess at medical diagnoses, create text-based Harry Potter games and explain scientific concepts at multiple levels of difficulty.

ChatGPT isn’t perfect, by any means. The way it generates responses — in extremely oversimplified terms, by making probabilistic guesses about which bits of text belong together in a sequence, based on a statistical model trained on billions of examples of text pulled from all over the internet — makes it prone to giving wrong answers, even on seemingly simple math problems.

There are also plenty of things ChatGPT won’t do, as a matter of principle. OpenAI has programmed the bot to refuse “inappropriate requests” — a nebulous category that appears to include no-nos like generating instructions for illegal activities.

Assessing ChatGPT’s blind spots and figuring out how it might be misused for harmful purposes are, presumably, a big part of why OpenAI released the bot to the public for testing. Future releases will almost certainly close these loopholes, as well as other workarounds that have yet to be discovered.

生词好句

1.brilliance

英 [ˈbrɪljəns] 美 [ˈbrɪljəns]
n. 才华，聪慧

2.weirdness

英 [ˈwɪədnəs] 美 [ˈwɪrdnəs]
n. 怪异，古怪

3.GPT

生成式预训练变换器（generative pre-trained transformer）

4.ever

英 [ˈevə] 美 [ˈevər]
adv. （进行比较时用以加强语气）以往任何时候，曾经
拓展:
It was my best ever score.
这是我有史以来最好的分数。
It was raining harder than ever.
当时下着前所未有的大雨。

5.release

英 [rɪˈliːs] 美 [rɪˈliːs]
v. 发布（to make a product available for the public）
拓展:
近义词：launch

6.breakthrough

英 [ˈbreɪkθruː] 美 [ˈbreɪkθruː]
n. 重大进展，突破；adj. 突破性的
拓展:
Scientists are hoping for a breakthrough in the search for a cure for cancer.
科学家们希望在治愈癌症的研究上取得突破。

7.come out

问世，推出；出来

8.cherry-pick

英 [ˈtʃeriˌpɪk] 美 [ˈtʃeriˌpɪk]
v. 挑选，精选，择优而取（to pick only the best people or things from a group）
拓展:
The new schools were accused of cherry-picking the best students in the area.
人们指责这些新学校只挑选这个地区里最好的学生。

9.not by any means

一点也不（by no means，not at all）
拓展:
ChatGPT isn’t perfect at all. / ChatGPT is by no means perfect.
ChatGPT并不完美。

10.oversimplify

英 [ˌəʊvəˈsɪmplɪfaɪ] 美 [ˌoʊvərˈsɪmpləfaɪ]
v. 过度简化
拓展:
simplify v. 简单化

11.pull

英 [pʊl] 美 [pʊl]
v. 获取（信息）；拉

12.be prone to …

易于遭受疾病的，有……（不好的）倾向的（likely to suffer from an illness or show a particular negative characteristic）
拓展:
I’ve always been prone to headaches.
我总是容易头疼。
He was prone to depression even as a teenager.
他在青少年时期 就很容易情绪抑郁。

13.as a matter of principle

出于原则

14.nebulous

英 [ˈnebjələs] 美 [ˈnebjələs]
adj. 模糊不清的，含糊的（not clear）

15.no-no

英 [ˈnəʊnəʊ] 美 [ˈnoʊnoʊ]
n. 不合适（或不可接受）的事物

16.blind spot

盲区，盲点

17.figure out

搞清楚，弄明白

18.misuse

英 [ˌmɪsˈjuːz] 美 [ˌmɪsˈjuːz]
v. 错用，滥用，误用

19.presumably

英 [prɪˈzjuːməbli] 美 [prɪˈzuːməbli]
adv. 很可能，大概，想必是（probably）

20.loophole

英 [ˈluːphəʊl] 美 [ˈluːphoʊl]
n. 漏洞
拓展:
close the loopholes 堵上漏洞
close the loop （形成）闭环

21.workaround

英 [ˈwɜːkəraʊnd] 美 [ˈwɜːrːkəraʊnd]
n. 变通办法，替代办法（a way of dealing with a problem or making something work despite the problem, without completely solving it）