跟读练习: Why AI won’t wipe out white-collar jobs | The Economist - 通过YouTube学习英语口语
C1
跟读控制
0% 已完成 (0/78 句)
Alex, what is AI going to do to our jobs?
⏸ 已暂停
速度:
重复次数:
等待模式:
字幕同步:0ms
所有句子
78 句
1
Alex, what is AI going to do to our jobs?
2
So of course, there's a lot of fear right now about what AI is going to do to the white-collar workforce.
3
And it's true that it could wreak havoc.
4
But I think it's much more likely to reshape jobs rather than to erase them altogether.
5
I think the proper way to think about this is that the future white-collar office is going to look less like a robot and more like a cyborg, where AI and humans are jointly working together.
6
We've heard so much about an AI apocalypse.
7
Where's your optimism coming from?
8
So I think it is useful to do two things.
9
First, it's useful to actually look at the data and see what's happened over the past three years.
10
And then second, it's useful to look at past technological revolutions and what happened during those.
11
So if you actually look at the data over the last three years and what happened to white collar jobs, you actually see that white collar employment has increased by three million jobs, whereas blue collar employment has stayed relatively flat over the first three years.
12
Certain occupations that have often been casted as the victims of AI, think software developers, think paralegals, think radiologists.
13
These are occupations that share a lot of overlap with the capabilities that AI can do.
14
And yet we've actually seen that these jobs too have seen massive employment increases over the last three years.
15
Software developers are up 7%, radiologists are up 10%, and paralegals have actually seen employment increase by 20% over the last three years.
16
So there's been large employment gains, but then wages too.
17
We see that white-collar workers continue to earn one-third more than blue-collar workers, which says that there is something distinct about office work that continues to command a wage premium over blue-collar work.
18
What has been the impact in the past of technological change?
19
With new technologies, there are always dire predictions for what the future of work is going to look like.
20
Even during the computer age, you had Nobel Prize economists predicting that the white-collar workforce was going to be completely eradicated in the coming decades.
21
Obviously, that's not what we've seen.
22
What we've seen instead is that computers, the Internet, have been a boon for white-collar workers since the early 1980s.
23
Employment in white-collar work has more than doubled.
24
Real wages, which are adjusted for inflation, have increased by a third for white-collar workers since the early 1980s.
25
And I don't want to oversimplify.
26
oversimplify.
27
There has been some job displacement because of computers.
28
Usually this is in routine work that is easily codifiable.
29
Think typists, which were completely wiped out.
30
But that has been completely overcompensated by two other effects.
31
The first is the reshaping of work.
32
So very rarely were full jobs completely amenable to automation.
33
And what you saw instead was that certain tasks would be automated and workers would then take on new roles that were actually higher productivity, higher value add.
34
Sort of a canonical example of this is air traffic controllers, where flight data could easily be automated by software.
35
And that allowed air traffic controllers to actually focus more on higher value-added activities such as judgment, such as coordination.
36
The second effect has been the creation of new work in new jobs altogether.
37
And this has been overwhelming since the early 1980s.
38
And it's important to emphasize that these new jobs do not just occur in roles that are quote-unquote computer roles.
39
Instead, what happens is the entire white-collar workforce is adjusted.
40
And so you have things like e-commerce, which spawns all types of new jobs in logistics, in supply chain coordination, in digital payments.
41
And you see new roles that you couldn't even think about 30 or 40 years ago.
42
Isn't it still possible that AI will be different?
43
Yeah, so the thinking goes, and where the fear comes in is that because AI is so much more powerful than past technologies, that it's going to slowly creep up the value chain into do ever-increasing complicated tasks.
44
And I think that this view does not hold weight for a few reasons.
45
First of all, AI, even with the impressive capabilities that it has today, it can still fully automate very, very few roles.
46
And so there was a report that came out by Anthropic late last year
47
that found that only in 4% of occupations could AI automate 75% of the job and that there were very few roles where AI could automate 100% of the job.
48
And so what this means is oftentimes we oversimplify what a job actually is.
49
Job consists of many different tasks.
50
Some of them will be automatable with AI.
51
But what this will do instead, similar to past technological revolutions, is it will shift the sort of work that people are doing.
52
So yes, data analysis, coding, a lot of these cognitive tasks will be able to be automated by AI.
53
And the most likely effect is that jobs will then continue to adjust, expand, and the sort of work that humans are doing will continue to shift.
54
And this is something that actually, if you look at the data, you're already starting to see in the labor market.
55
So what does that data show?
56
We found that a lot of the jobs that combined both technical work and human skills think roles that involve a lot of coordination, such as project managers, roles such as information security analysts.
57
These occupations have actually seen employment expand by more than 30 percent over the last three years.
58
So there is a subset of occupations that are really, really expanding rapidly.
59
The only group of occupations that we actually saw shrink over the last three years were occupations that could be considered routine back office roles.
60
Secretaries and administrative assistants, for example, have seen employment shrink by 20% over the last three years.
61
This is, of course, a concern, but we see it more as a continuation of a past pattern rather than sort of a structural shift because of AI.
62
Alex, does this mean we can stop worrying about the destruction of white collar jobs?
63
So I know I've been painting a somewhat optimistic picture.
64
I think that there are certainly risks as well that are important to acknowledge.
65
The first is that models are becoming ever more capable.
66
There are a lot of benchmarks that show the capabilities of models and the most recent models by Anthropic, their cloud models, can now do software tasks continuously for more than five hours.
67
the rate at which these models are increasing over the past couple of years has doubled every seven months.
68
So there is this one concern is that the models just continue to get better and better.
69
The second concern is that entry-level jobs are the ones that are most highly vulnerable and highly exposed to what AI can do.
70
So interns, entry-level analysts, entry-level programmers could see
71
possible disruption and thinking through what this means for career ladders for workers who are just joining the labor force is something that should be taken seriously.
72
And then the third concern is that a lot of the jobs that we have seen employment decrease the most over the past three years, such as these back office occupations, are also the workers who are least adaptable
73
to finding new work.
74
And so a lot of thinking needs to be done in how to help workers that do lose their job because of AI, how to help them transition into new roles.
75
But with that being said, overall, I think despite all of the sort of dire predictions and all of the calls for a white collar jobs apocalypse,
76
I do believe that white collar work will continue to adjust and most probably it will continue to expand as well.
77
Alex, thank you so much.
78
Thank you, Rosie.
📱
Shadowing English
现已推出移动版，立即下载！
5.0
为何要通过此视频练习口语？
通过观看和模仿这段视频，你可以在真实的语言环境中提高英语口语能力。这段对话讨论了人工智能对白领工作的影响，涵盖了多个相关主题，提供了丰富的词汇和表达方式。练习时，建议使用英语影子跟读（shadow speak）的方法，即在听的同时重复发音，这样可以帮助你更好地掌握语言的节奏和语调，加深对语句结构的理解，通过英语口语练习提高自信心。
语法和表达方式分析
在视频中，出现了多个重要的语法结构和表达方式，以下是一些示例：
- 被动语态（被预测 / 是否会被消除）— 此结构常用于强调动作而非执行者，如“白领工作将不会被彻底消除”。
- 条件句（如果......那么）—用于表达假设性的情况，例如：“如果AI真的能替代职能，那么...” 这帮助听者理解可能的结果。
- 对比结构（白领与蓝领的比较）— 用于突出两个对象的不同，如“白领收入比蓝领高出三分之一。”通过这些结构，你可以学习如何在对话中有效地建立对比。
常见发音陷阱
在学习过程中，准确的发音是关键。在此次视频中，以下几个词汇可能会成为发音的难点：
- automation（自动化）— 容易读成“自动马”，而实际发音应注意重音位置和音节分割。
- roles（角色）— 此词在快速对话中可能会被忽略部分音节，建议慢速练习，确保每个音节清晰可闻。
- technology（技术）— 发音时需注意“n”和“g”的音，确保清晰连贯。
通过不断的提高英语发音，你可以减少沟通中的误解与障碍。练习时，建议反复听录音并模仿发音，同时尝试整合到你的日常对话中。
什么是跟读法？
跟读法 (Shadowing) 是一种有科学依据的语言学习技巧，最初开发用于专业口译员的培训，并由多语言者Alexander Arguelles博士普及。这个方法简单而强大：您在听英语母语原声的同时立即大声重复——就像是一个延迟1-2秒紧跟说话者的影子。与被动听力或语法练习不同，跟读法强迫您的大脑和口腔肌肉同时处理并模仿真实的讲话模式。研究表明它能显着提高发音准确性，语调，节奏，连读，听力理解和口语流利度——使其成为雅思口语备考和真实英语交流最有效的方法之一。