哪些專業可能會被 AI 所取代？
在機器人、人工智慧發展的同時，依舊有很多問題困擾著我們，就像在《你說的都對》節目中，知識大神淡豹認為，未來很多高薪工作會被人工智慧所取代，像是律師行業，你認同她的說法嗎？哪些專業可能會被AI所取代？你認為以後人類將會做什麼行業？
謝謝邀請。富有創造力和藝術性的行業，AI很難大施拳腳。純體力活、重複勞動、更新變化慢的工作容易受AI影響。之前看到一篇英文文章，講的就是哪些工作最容易被AI取代，哪些工作最不容易受AI取代，小Q在此分享一下，為保留作者原意，本處不作中文翻譯。
10 Careers AI Will Replace (and 10 That Are Safe)
Most Likely to Be Replaced
1) Telemarketers
Likelihood: 99%
Why: You probably already receive robo-calls on behalf of various products and services, and career growth in the telemarketing space is expected to decline by 3% by the year 2024. This is largely in part because of the requirements to be successful: Unlike other sales roles, telemarketers dont require a high level of social, or emotional, intelligence to be successful. Think about it -- are you likely to purchase from a telemarketer? Conversion rates for direct telephone sales are typically less than 10%, making this role a ripe opportunity to be automated.
2) Bookkeeping clerks
Likelihood: 98%
Why: Jobs in this role are expected to decline 8% by 2024, and its no surprise why -- most bookkeeping is becoming automated, if it hasnt been already. QuickBooks, FreshBooks, and Microsoft Office already offer software that does the bookkeeping for you thats much more affordable than a persons salary, so its no surprise this job has such a high probability.
3) Compensation and Benefits Managers
Likelihood: 96%
Why: This one is surprising because the job growth is supposed to increase 7% by 2024. But just because theres demand doesnt make you safe from automation. As companies grow in size -- especially across multinational markets -- a human and paper-based system can present more hurdles, time delays, and costs. Automated benefits systems can save time and effort for providing benefits to large numbers of employees, and companies like Ultipro and Workday are already being widely adopted.
4) Receptionists
Likelihood: 96%
Why: Pam predicted this back on The Office, but in case youre not a fan, automated phone and scheduling systems can replace a lot of the traditional receptionist role -- especially at modern technology companies that dont have office-wide phone systems or multinational corporations.
5) Couriers
Likelihood: 94%
Why: Couriers and delivery people are already being replaced by drones and robots, so its only a matter of time until this space is dominated by automation altogether. At the same time, this space is expected to grow by 5% by 2024, so it might not happen as quickly as you think.
6) Proofreaders
Likelihood: 84%
Why: Proofreading software is everywhere -- and we use it a lot here at HubSpot. From Microsoft Words simple spelling and grammar check to Grammarly and Hemingway App, there are a lot of technologies out there that make it easy to self-check your own writing.
7) Computer Support Specialists
Likelihood: 65%
Why: The field is projected to grow 12% by 2024, but with so much content on the internet with instructions, step-by-step guides, and hacks out there, its no surprise companies will rely more heavily on bots and automation to answer support questions from employees and customers in the future.
8) Market Research Analysts
Likelihood: 61%
Why: Market research analysts play an incredibly important role in the development of messaging, content, and products, but automated AI and surveys can compile this information more and more easily. GrowthBot, for example, can conduct market research on nearby businesses and competitors with a simple Slack command.
9) Advertising Salespeople
Likelihood: 54%
Why: As advertising shifts away from print and TV and towards web and social media landscapes, people simply dont need to be managing those sales for marketers who want to buy ad space. More social media platforms are making it easy for people to buy space through free application program interfaces (APIs) and self-serve ad marketplaces to remove the salesperson and make it faster and easier for users to make money -- and thats reflected in the projected 3% decline in the industry.
10) Retail Salespeople
Likelihood: 92%
Why: If youve visited a mall, car dealership, or furniture store lately, you might not have been assisted by a salesperson at all from start to finish. Companies are democratizing the shopping experience with features like self-checkout, and the modern buyer is much more internet-savvy and more likely to do internet research and make a buying decision on their own.
Most Likely to Be Safe (For Now)
1) Human Resources Managers
Likelihood: 0.55%
Why Not: Its kind of in the name -- but your companys Human Resources department will likely always need a human at the helm to manage interpersonal conflict with the help of non-cognitive and reasoning skills. The field is projected to grow 9% by 2024 as companies grow and need more robust structures for supporting and helping employees.
2) Sales Managers
Likelihood: 1.3%
Why Not: Sales managers need a high level of emotional intelligence to hit their quotas each month, network and collaborate with customers, and motivate and encourage the larger sales team. Managers also have to analyze data and interpret trends, and the high levels of intelligence required -- plus the constant need to adapt to new situations -- makes this role safe from automation.
3) Marketing Managers
Likelihood: 1.4%
Why Not: Marketing managers have to interpret data, monitor trends, oversee campaigns, and create content. They also have to nimbly adapt and respond to changes and feedback from the rest of the company and customers, making this another human-forward career AI isnt quite ready to replicate.
4) Public Relations Managers
Likelihood: 1.5%
Why Not: Successful PR managers rely on a network of relationships and contacts to procure press placements and buzz for the companies they represent, making this another completely safe role. PR managers who have to raise awareness around an issue or mission need a particularly human touch to raise funds or get people to participate in a campaign, too -- and jobs are expected to grow 7% by 2024.
5) Chief Executives
Likelihood: 1.5%
Why Not: Its nearly impossible to automate leadership -- after all, its hard enough to teach it. Chief executives have to inform broad strategy, represent companies missions and objectives, and motivate huge teams of people working for them. Companies may answer to stakeholders and boards of directors, who likely wouldnt want a robot giving them an earnings report, either.
6) Event Planners
Likelihood: 3.7%
Why Not: Event planning is a growing field, and if you ask anyone on our events team here at HubSpot, whether youre planning an event for employees, customers, or an industry event with tens of thousands of attendees, the planning process has many, many moving parts involved. Planners have to coordinate and negotiate with vendors, contractors, and freelancers to make things come together, and the organizational and people skills involved will make this another near-impossible role to automate.
7) Writers
Likelihood: 3.8%
Why Not: (I breathed a sigh of relief on this one.) Writers have to ideate, create, and produce original written material. AIs can do some of this with title suggestions, writing prompts, and automated social media messages, but blog posts, books, movies, and plays will likely be written by humans for the foreseeable future.
8) Software Developers
Likelihood: 4.2%
Why Not: Software engineering and development is hard enough for human beings to do, and the time and skill investment needed to create applications, software, and websites will be tough to replicate -- especially since developers need to execute perfectly to create great products for customers. The field is expected to grow by 19% by 2024, so if youre a software developer, youre sitting pretty for now.
9) Editors
Likelihood: 5.5%
Why Not: While some of the load can be lifted from editors with the automated proofreading technology mentioned previously, editors have to review writers submission for clarity, accuracy, comprehensiveness, and originality. While there is some software that can spot-check for clarity and scan for plagiarism, the editor role must be carried out by a human in order to read work as another human would.
10) Graphic Designers
Likelihood: 8.2%
Why Not: Although there are some AIs taking small (and somewhat creepy) steps in the graphic design space, graphic design is both artistic and technical, making it an ideal role for a human being to carry out. Like writing, all work needs to be original and created to the clients wishes, so graphic design needs to be created with a human artist and editor all-in-one.
From: 10 Jobs Artificial Intelligence Will Replace (and 10 That Are Safe)
謝邀。我之前寫過幾篇關於這個問題的文章，摘錄一些段落：
淺薄：通向被機器取代之路 | 透明思考?gigix.thoughtworkers.org
阿蘭·圖靈於1936年發表的論文不僅奠定了現代計算機的理論基礎，同時也指出了計算機的邊界：這個世界上有無窮多的問題是計算機不能解決的。這一發現在哲學層面的意義絲毫不亞於其使用價值，因為它實際上是在論斷：人的思想無法被計算機取代。
不過嚴格說起來，這個論斷的成立還需要另一個前提條件，即那些計算機不可計算的問題，人腦是可以解決至少其中一部分的。而尼古拉斯·卡爾在《淺薄：互聯網如何毒化了我們的大腦》中指出，這個前提條件正在隨著計算機和互聯網的飛速發展而變得岌岌可危。計算機仍然不能解決那些不可被計算的問題，然而我們自己卻在越來越多地使用計算機的過程中逐漸失去人類獨有的那部分思想優勢。
正如布林約爾松和麥卡菲在《與機器賽跑》中所說，在一個計算機能力日益強大的時代，人類獨特的競爭力在於情感交流與創新。然而適應於時代的、碎片化的信息輸入方式不僅會使我們的思想缺乏深度而失去創造性，而且讓我們失去深入的主觀情感判斷，甚至陷入道德相對主義與道德虛無主義。於是我們看到一個自我增強的困境：機器不僅在取代人類用淺薄的思想可以做的那些工作，而且還將更多人的思想變得越來越淺薄。馬克思說勞動者異化成勞動的工具，而這一次的異化更可怕之處在於：淺薄的思想無法意識到自身被異化。難道我們真的已經無法抗爭了嗎？
人與機器的終極競賽 | 透明思考?gigix.thoughtworkers.org
美國的人口在過去十年間增加了3000萬，但就業機會幾乎沒有增加，於是就業率從2000年的64%跌至不到58%。如前所述，這次的失業是結構性的。造成失業的原因不是增加裁員，而是缺乏僱傭。整個工業時代一以貫之的僱傭與GDP的正相關性，從此一去不復返了。
在這場技術造成的結構性失業中，我們可以看到三組「勝利者 vs. 失敗者」的對比：（1）高技術工人 vs. 低技術工人；（2）超級明星 vs. 平凡大眾；（3）資本 vs.勞動力。而且這三組勝利者並非互斥：其中一方面的成功者有更大幾率在另外兩方面也取得成功。於是財富更多地向少數人聚攏。
新技術一方面可以自動化重複性勞動（例如機器人、數控機牀、數字化倉儲控制），使低技術工人失業；另一方面又增加交流、聯想和創造的能力（例如數據可視化、數據分析、高速通訊、快速原型），提升高技術工人的價值。這場技術革命是一邊倒地有利於高技能人才的，因此被稱為「技能偏向的技術變革」。儘管各國都在強調教育，但對技能的供給遠遠跟不上需求。
有趣的是，受傷害最大的不是最低技術的純物理服務者（例如餐廳服務員），而是中低技術水平的知識/技術工作者（例如銀行櫃員、電話客服）。莫拉維克悖論指出，邏輯推理只需要很少的計算能力。這也就使得這些基本能夠以符合邏輯的流程來定義的工作更容易被機器所取代；反而是一些在人類看來簡單的事情，例如在複雜地形上行走、識別人臉、天馬行空地閑聊，是機器很難做到的。
但所謂「只有人能做的事」這個疆域在不斷地受到挑戰。Google的汽車自動行駛一萬英里無事故；IBM的Watson參加百科知識答題秀贏過了所有人；eDiscovery取代了律師的眾多日常工作；醫生的診斷也正在被機器取代。我們可以看到一些事情是機器更難做到的：（1）複雜的模式匹配；（2）複雜的交流。但是沒有任何證據表明機器不能做到這些事。
鮮為人知的，現在最強的國際象棋棋手不是某臺計算機（當然更不是某個人），而是一支兩個業餘棋手與三臺計算機的隊伍。有趣的是，很強的職業棋手和計算機組隊並沒有獲勝。因為傳統意義上棋手的「強」，有很大部分是可以被計算機取代的。組隊的實力關鍵不在於人類棋手的計算能力，而在於人類棋手與計算機的配合。同樣的模式出現在很多領域：醫學、法律、財務、零售、製造、甚至科學探索。人類的價值不是體現在和機器面對面競賽，而是體現在與機器搭檔競賽，使用那些機器所不擅長的技能：複雜的模式匹配，複雜的交流，創新。
創新來自於對以往創新的組合與重組。瞭解的人和思想越多，產生新點子的可能性就越大。因為創新是一種組合與重組，所以創新的可能性是指數性的。這意味著兩件重要的事：第一，我們不需要擔心創新被用盡；第二，「從所有可能性中發現有價值的點子」這件事是NP不可解的，也就是計算機從本質上不擅長的。
書評：與機器賽跑 | 透明思考?gigix.thoughtworkers.org
兩位作者指出，人類的特長在於複雜的、涉及情感的交流，以及複雜的模式匹配和創造性的思維。然而與一些樂觀主義者不同，兩位作者並不認為人類自然而然地就會具備與計算機協作所需的那些能力——從日常生活中我們就能看到，信息碎片化、深度閱讀與思考的失位、電腦和手機對人際交往的侵入……信息技術不僅沒有增強、反而在潛移默化地弱化人們進行複雜交流和創造性思維的能力。陰謀論者簡直可以就此寫一篇科幻小說，談論覺醒的硅基生命如何逐步奴役人類了。
在較短時間內讓絕大多數普通人具備與機器協作的能力，教育將扮演至關重要的角色，因此兩位作者也表達了他們對現行教育體制的憂慮。
首先，教育必須給人們傳授那些非機械化的、非重複性的、涉及情感與創造性的技能，因為只有這些技能纔是不可被機器取代、因此在下一個時代仍然有效的。以中國人常引以為豪的數學教育為例，現在學生們從課堂上學到的多是「執行演算法」的能力而非「設計演算法」的能力，而前者恰好是計算機的長項，後者纔是人類可能與計算機展開合作的領域。諸如此類的教育思路與方式若不及時改變，在一兩代人之後人類被機器廣泛取代絕非危言聳聽。
基於當前技術的發展程度與合理推測，在15年內，人工智慧和自動化將具備取代40%~50％崗位的技術能力。
這些崗位主要集中在以下工作和任務場景：
重複性勞動——特別是在相同或非常相似的地方完成的工作，如洗碗、裝配線檢查、縫紉；
有固定臺本和對白內容的各種互動，如客戶服務、電話營銷；
相對簡單的數據分類，或思考不到1分鐘就可以完成識別的工作，如文件歸檔、作業打分、名片篩選；
在某公司一個非常狹小的領域工作，如銀行理財產品的電話推銷員、某部門的會計；
不需與人進行大量面對面交流的工作，如分揀、裝配、數據輸入。
機器人進展比AI軟體慢：它們仍然非常笨拙
儘管AI大幅度提升生產效率，節省人力資本，但為了確保人類的職業生涯不會因人工智慧替代而中斷，需要了解「在可見的未來裏，人工智慧做不到什麼」。
雖然有媒體報道，稱巨額投資將用於開發人工智慧和機器人，但自動駕駛汽車、人工智慧放射科醫師等人工智慧應用可能需要很長時間才能普及。
因而有幾類工作難以取代：
創意性工作都，例如醫學研究員、人工智慧科學家、獲獎劇本作家、公關專家、企業家；
複雜性/戰略性工作，例如首席執行官、談判專家、併購專家；
靈敏性工作，例如口腔外科醫生、飛機機械師、脊椎按摩師。
應該鼓勵批判性思維和各種創造力，不僅僅是科學和工程，還有藝術、建築、音樂、詩歌、表演、講故事等。應該接受傳統工作崗位正在流失的現實。對於老年人，當需要提前退休時，請考慮接受，通過打零工和志願服務賺取一些收入並過上喜歡的生活。
謝邀，我較為認同她的說話。一些繁重、重複、能用機器完成的工作，正在或後期都將被機器取代，比如翻譯、客服、律師等，當然這也不是一個一蹴而就的過程，技術的發展有周期，那麼職業的發展也是類似的；
未來有多遠？長期有多長？人工智慧足夠發達時候，未來所有專業都會被替代，長期來看人類都會被替代。
作為一個喫貨，先從喫來講。
食物是從地裏種出來的，是飼養起來的（捕殺的比較少。。就暫時忽略了）。這種農業與畜牧業生產都可以被全自動化，輔佐以智能的演算法，養養豬什麼的就不勞煩老丁了。
再到廚藝，人工智慧做這個也是很有天賦的，再厲害的廚師也做不到辣麼精準的味覺把控。無論高級餐飲還是尋常家菜，只要你有原料，有能源，理論上就能在家享受米其林餐廳大廚的手藝。所以，農夫與廚子還有屠夫是註定要涼涼的。
想喫到多種好喫的必然離不開發達的物流，物流就離不開司機，而無人駕駛實現24小時高效安全運輸非常容易取代人工司機。至於快遞，只要各個區域有一個取貨站，無人配送就很容易了（不能爬樓送到家門口）。所以，到了這，大部分司機與快遞員包括外賣小哥可能也涼涼了。
而有的餐廳總是那麼「異類」，你需要親自過去才能喫，走得遠就離不開城市的交通，人工智慧管理的城市交通網與無人駕駛可以幫助你以最快的速度且零拖延（拒絕等紅燈）把你安全送到目的地。如果你想去的餐廳在別的城市，你甚至可以體驗到在不停移動中的火車站進入高鐵，整個過程包括安檢都不會對您造成阻礙（不要攜帶危險品乘車～）。如果您需要乘坐飛機，可以選擇無人駕駛的客機將您平穩送達。到這，交通管理的，公交車計程車司機，以及安檢，飛行員通通都是可替代的。
每個人都希望在優雅的環境中用餐，而24小時維持餐廳的各項要求交給人工智慧與機器人再合適不過了。您的餐桌總是能滿足您各項需求。所以，服務行業的大部分工作都涼了。
而AI無法取代的是什麼呢？或許只是一個想法，一個咬一口蘋果的想法。（隨隨便便答的，其實大部分都已經實現了）未來只有三種職業！1，老師！2，程序員！3，維修員！
推薦閱讀：
