Are You Prepared to Be Interviewed by an AI?

在疫情隔離期間，許多人力資源管理程序轉(zhuǎn)移到線上。即使生活日漸重歸親臨實地的現(xiàn)實，這些程序中有許多也會繼續(xù)保留。其中之一是虛擬的，或者說預先錄制的求職面試，它們越來越多地受人工智能（AI）引導。這些面試通常會縮短招聘過程，讓企業(yè)以更低廉的成本找到合適的候選人。

During the isolation of the pandemic,?many human resource management processes moved online. And even as life goes back to being more in person, many of these processes are here to stay. One is virtual, or pre-recorded job interviews, which are increasingly guided by artificial intelligence (AI).?These interviews typically shorten the hiring process, making it cheaper for companies to find the right candidates.

這種轉(zhuǎn)變永遠改變了求職者的體驗，而且并非總是向好的方向改變。今天，年輕的求職者在尋找他們的第一個職位、見習或?qū)嵙暪ぷ鲿r，很可能在他們的第一次面試中面對的是一個機器人，而不是人。

This shift has forever changed the?experience of jobseekers, and not always for the better. Today, younger job seekers looking for their first role, placement, or internship are likely to face a bot at their first interview, not a human.

在最極端的自動化視頻面試（automated video interviews, AVIs）類型中，機器人會提出幾個預先確定好的問題，給求職者短暫的時間來回答這些問題，并當即對這個人做出判定。我們將這些定義為AI主導的面試。

And in the most extreme type of automated video interviews (AVIs), a bot asks a few predefined questions, giving the candidate a short window of time to answer them, and makes a decision about the person right then and there. We define these as AI-led interviews.

雖然自動化視頻面試與傳統(tǒng)的當面面試或電話面試的某些方面保持未變（求職者仍然必須給人留下良好的第一印象），但對于求職者而言，AI主導的面試仍有許多不為人知的地方。這份簡短的指南提供了駕馭這些問題的一些建議。

While some things about AVIs and traditional in-person or phone interviews remain the same (job candidates?must still make a good first impression), there’s much that remains mysterious to job seekers about AI-led interviews.?This short guide provides some advice on navigating these questions.

何為自動化視頻面試？

What Are Automated Video Interviews?

準備面試的第一步是要知道你要參加的是哪種面試。根據(jù)技術(shù)在選拔過程中的使用方式，面試有各種不同的類型，而AI可以不同程度地參與其中。

The first step to preparing is?becoming aware?of what kind of interview you are about to take. There are different types, depending on how technology is used in the selection process, and AI can be involved to varying degrees.

同樣重要的是，要注意，并非所有的視頻面試都會使用AI技術(shù)。比如，有些面試只是在錄制視頻，然后交由招聘經(jīng)理觀看。其他一些面試會讓AI參與處理視頻錄制期間收集的不同類型的數(shù)據(jù)。

It is also important to note that not all video interviews use AI technology. For example, some interviews are just a recording of a video, which will then be watched by hiring managers. Others will involve AI processing different types of data collected during the video.?

為了確認你即將參加的面試是否真的包括AI，請密切留意你收到的有關(guān)面試的電子郵件中或者精美印刷郵件中是否有以下詞語：機器學習、預測分析、決策算法、推薦引擎或數(shù)據(jù)驅(qū)動的決策。

To be sure your upcoming interview indeed includes AI, keep an eye out for the following terms in the emails you receive about the interview, or in the fine print:?machine learning, predictive analytics, decision algorithms, recommendation engines,?or?data driven-decision.

如果你確實發(fā)現(xiàn)了這些詞語，你可以預期可能有三種類型的數(shù)據(jù)會被收集：視覺數(shù)據(jù)（面部表情、眼部動作、手部動作）、語言數(shù)據(jù)（詞匯、關(guān)鍵詞）和聲音數(shù)據(jù)（語音語調(diào)、發(fā)音）。AI主導的面試使用收集到的數(shù)據(jù)自動生成預測，看看候選人是否他們正在物色的人選。比如，旨在預測你是否是一位優(yōu)秀候選人的AI可能會使用過去錄用人員及其隨后工作表現(xiàn)方面的大數(shù)據(jù)，從而預測可能與更優(yōu)秀工作表現(xiàn)相關(guān)的候選人特征。

If you do identify these terms, you can expect that three types of data might be collected: visual (facial expression, eye movement, hand movement), verbal (vocabulary, key words), and vocal (voice tone, pronunciation). AI-led interviews use the data collected to automatically generate a prediction on whether the candidate is the person they are looking for. For example, AI designed to predict whether you are a good candidate or not might use big data from past hires and their subsequent performance at work, predicting candidate characteristics that might correlate to higher work performance.

AI主導的面試并非超群的選拔工具（到目前為止）

AI-Led Interviews Are Not Superior Selection Tools (Yet)

使用AI的面試具有一些共同之處：你會發(fā)現(xiàn)自己在一塊屏幕前，有時你會在屏幕里看到自己的影像，回答著自動提出的問題，幾乎沒有時間思考。人與人之間的互動以及我們從面試官那里得到的實時提示不見了蹤影，而那些提示可以讓我們立即感覺到自己的表現(xiàn)如何。到辦公室或其他實體建筑的過渡時間也不復存在；不再有前往未來可能工作地點的旅程，不再有環(huán)境的改變，讓你可以做好準備，在心理上切換到面試模式。

Interviews using AI have something in common: You will find yourself in front of a screen, sometimes with your own image reflected back at you, answering automated questions with little time to think. Gone are the human interactions and real-time cues we get from our interviewer, which give us an immediate sense of how we are doing. Also gone are the transition times to the office or other physical building; there is no more journey to prospective work locations, no shift of an environment that prepares you to mentally switch to interview-mode.

在我們的研究中，我們發(fā)現(xiàn)，這種體驗通常讓人困惑、令人不安。求職者被拽向兩個方向：

一方面，這項技術(shù)的新穎性和有時被認為是AI使然的“卓越”品質(zhì)讓他們覺得這是一種未來主義的體驗，必然會克服人類的偏見。

另一方面，面試過程中人情紐帶的缺乏以及當時的緊張氣氛交融在一起，令人生畏，加劇了焦慮。我們要告誡的是，對技術(shù)的美化通常是基于對AI能（和不能）做之事的理想化和了解不足。

In our?research, we found that this experience is often confusing and unsettling. Job seekers are pulled in two directions:

On one hand, the novelty of the technology and the “superior” quality that is sometimes attributed to AI makes them feel this is a futuristic experience bound to overcome human bias.

On the other, the lack of human connection during the interview and the tension of the moment is a daunting mix that heightens anxiety. We warn that the glorification of the technology is often based on an idealization and a poor understanding of what AI can (and can’t) do.

我們發(fā)現(xiàn)，候選人美化技術(shù)的傾向使他們相信，技術(shù)會做出比人類更好的決定。這導致他們感到被某種更高級的實體“評判”。我們則相反，鼓勵候選人認識到，這項技術(shù)根本不完美。相反，這些類型的面試可能效度欠佳；比如，它們并不擅長根據(jù)視頻中提取的言語和非言語行為來預測個性特征。這意味著該技術(shù)通常不足以衡量它本該衡量的東西，或者換句話說，這項AI并不像宣傳的那樣先進。難怪，一些分析家認為人臉識別和其他技術(shù)的使用是偽科學，一些法院已將其列為非法技術(shù)。

We found that candidates’ tendency to glorify the technology made them trust it would make?better decisions?than human ones. This resulted in them feeling “judged” by a sort of superior entity. We instead encourage candidates to understand that the technology isn’t perfect by a long shot. Instead, these types of interviews may suffer from?poor validity; for example, they are not very good at predicting personality traits from verbal and non-verbal behaviors extracted from videos. That means that the technology is often not good enough to measure what it is supposed to measure or, in other words, that the AI is not as advanced as advertised. Unsurprisingly, some analysts have deemed the use of facial recognition and other technology as?pseudoscience, and some courts have?outlawed?it.

因此，我們建議以務實的態(tài)度看待該技術(shù)做出的任何判斷，要認識到這樣的判斷不精確，存在潛在缺陷和偏見。換言之，在面試前或面試期間，不要讓AI的判斷擊倒你的自信心！

So, we advise taking a pragmatic view of whatever judgement is made by the technology as imprecise and potentially flawed and biased. In other words, do not let AI judgment knock down your self-confidence prior to, or during, an interview!

給人留下好印象，意味著要保留人情味

Making a Good Impression Means Staying Human

盡管自動化視頻面試存在缺陷，特別是那些利用AI的自動化視頻面試，但你很可能還是會遇到它們。那么，你如何才能最好地應付這些面試呢？許多AI面試平臺建議你只需“做你自己”。然而，在我們之前的研究中，我們注意到，這恰好是求職者感覺有難度的地方。面對自動化視頻面試，他們通常舉止不自然。人們向我們講述過，他們?nèi)绾我宰哉J為會更好看的方式來調(diào)整自己，保持僵硬的姿勢，目不轉(zhuǎn)睛，并盡量少用自己的手。為了給算法留下好印象，我們的許多受訪者向我們表示，他們感覺自己正在變成機器人。

Despite the flaws of AVIs, particularly those that utilize AI, it’s likely you’ll still encounter them. So how can you best approach these interviews? Many AI interview platforms suggest that you just?“be yourself.”?And yet in our?previous research, we noticed that this is precisely what job candidates find difficult. Faced with an AVI, they tended to behave in?unnatural ways. People told us of how they adjusted themselves in ways they thought would make them look better, keeping a rigid posture, a fixed gaze, and using their hands as little as possible. In trying to make a good impression on the algorithm, many of our interviewees told us they felt like they were becoming robots themselves.

給人留下良好第一印象的關(guān)鍵是，不僅要了解所涉及的技術(shù)，而且要了解在技術(shù)面前你可能承受的壓力。其訣竅是盡量保持自然——盡管這可能會讓人感覺很不自然。因此，為了抵制機器人般僵硬的本能反應，我們建議接受面試的人應該進行練習，首先是在有其他人在線上出席的情況下，然后是單獨練習。

Understanding not just what technology is involved, but also the pressure you might be under in the presence of the technology, is key to making a good first impression. The trick is to stay as natural as possible — despite how unnatural this may feel. So, to counter the knee-jerk reaction of robot-like rigidity, we suggest that interviewees should practice, first in the presence of other humans online, and then solo.

第一步，要習慣于對著屏幕說話。請一個朋友使用Zoom或WhatsApp視頻，讓他們向你提出預先安排好的問題。面試問題越來越多地可以在不同大學網(wǎng)站提供的題庫中找到，這些網(wǎng)站是讓學生為面對這些面試做準備的。

First, get used to speaking to a screen. Ask a friend to use Zoom or WhatsApp Video and have them ask you prearranged questions. Increasingly, interview questions can be found in?question banks?offered by different university sites that?prepare?their students to face these interviews.

我們建議采取三步走的方法來進行這種角色扮演。首先，通過打開攝像頭進行練習，這樣當你在屏幕上看著你的朋友們時，他們可以向你提問。起初，另一個人的存在會讓你感到安心，并幫助你找到回答問題的信心，就像你在正常的面試中那樣。你要給自己錄像，回放，并分析你做得好的地方。要記住，積極心理學告訴我們，關(guān)注長處，而不是關(guān)注發(fā)展點，可以加速改善理想中的行為。

We suggest a three-step approach to this roleplay. Initially, practice by having your cameras on, so that your friends ask questions while you see them on screen. Initially, the presence of another human will be reassuring and help you find the confidence to answer, as you would in a normal interview. Record yourself, play it back, and analyze what you did well. Remember,?positive psychology?tells us that focusing on strengths, rather than on development points, can result in a faster improvement of desired behaviors.

第二步，重復同樣的練習，讓你的朋友提問，但要把攝像頭關(guān)掉。對著黑屏說話會更讓人不自在，可是，你會越來越接近于使用自動化視頻面試的情形。還是老樣子，要給你自己錄像，分析錄制的內(nèi)容，找出你做得好的地方，并注意這一輪是否存在任何差異。

Second, repeat the same exercise, with your friend asking questions, but with the camera off. It will be more awkward to speak to a black screen, but you will be getting closer to what it will be like to use AVI. Again, record yourself, analyze the recording, identify what you did well, and note whether there were any differences this round.

第三步，在一份文件中準備幾個問題，然后獨自練習。對著你的電腦屏幕說話，并給自己錄像。你做得好的方面有哪些？你的哪些行為與你面對人類時有所不同？通過分析這三個步驟，你可以意識到自己在不同類型的條件下表現(xiàn)如何，而且通過練習，你的自然舉動相比于人與人的對話會毫不遜色，哪怕是在自動化視頻面試期間。

In the third step, prepare a few questions in a document and go solo. Speak to your computer screen and record yourself. What did you do well? And what did you do differently than when you were facing a human? By proceeding through all three steps analytically, you can become aware of how you perform under different types of conditions, and with practice, you’ll be able to match your spontaneity to a human-to-human conversation, even during an AVI.

最后，雖然心理準備是為自動化視頻面試做準備的一部分，但實踐部分也很重要。我們發(fā)現(xiàn)，在我們的研究中，成功的候選人花時間確保他們的環(huán)境有點“工作性質(zhì)”。他們有素凈的背景，確保燈光和聲音良好，以便他們在鏡頭前可以得到良好的拍攝效果，并反復檢查他們的技術(shù)是否正常工作。在面試前，他們還在這些相同的環(huán)境中進行排練，并且留出足夠的時間來舒適地面試。從本質(zhì)上講，他們的行為就好像他們必須經(jīng)過心理上和身體上的準備才能參加當面面試一樣。

Finally, while the psychological preparation is part of getting ready for an AVI, the practical part is important, too. We found that the successful candidates in our research spent time making sure their environment was somewhat “work-like.” They had a neutral background, ensured that the lighting and sound was good so they could be captured well on camera, and double-checked that their technology was working properly. They also did their rehearsals in these same settings before the interview, and blocked out enough time to do the interview comfortably. Essentially, they behaved as if they had to go to an in-person interview by preparing both psychologically and physically.

作為一種新的招聘形式，自動化視頻面試對求職者而言可能具有陡峭的學習曲線。我們要求你對這個過程和你的表現(xiàn)進行反思，而不要評判自己。我們是人，而在如此重要的時刻——求職面試——不得不面對一種新技術(shù)，這會產(chǎn)生不適和焦慮。可是，就像其他任何事情一樣，你對自動化面試的經(jīng)驗越豐富，你的表現(xiàn)就越優(yōu)秀。

AVIs, as a novel form of recruitment, can have steep learning curve for job candidates. We ask you to be reflexive of the process and of your performance without judging yourself. We are humans, and having to face a new technology at such an important moment — a job interview — can create discomfort and anxiety. But like everything else, the more experienced you get with automated interviews, the better you’ll perform.

要記?。哼@是一項新技術(shù)。許多研究告訴我們，它并不像你想象的那么聰明，而且在精確了解你是何人方面也十分有限。所以不要讓它動搖你的自信。

And remember: The technology is new. A lot of research tells us that it is not as clever as you might think and is limited in understanding exactly who you are. So don’t let it rattle your self-confidence.

扎希拉·亞澤是哲學博士，蘇塞克斯大學商學院的副教授，蘇塞克斯大學MBA的主任，以及工作中的數(shù)字未來研究中心的副研究員。

迪米特拉·彼得拉卡基是蘇塞克斯大學商學院的技術(shù)與組織學教授，也是由經(jīng)濟和社會研究理事會（ESRC）資助的工作中的數(shù)字未來研究中心的共同投資人。

時青靖 | 編輯