跳至主要内容

第一份「生成式人工智慧證書」來了，附參考答案

課程介紹​

半個月前，微軟和領英合作推出了第一個「生成式人工智慧證書」，career-essentials-in-generative-ai-by-microsoft-and-linkedin，這是個免費的有包含生成式AI、人工智慧原理、AI倫理等五節內容的免費課程，完成學習並且透過考試後就能得到對應證書了（如題圖），可以把證書加到自己的領英頁面，也可以寫在簡歷裡。

這個證書不能說含金量有多高，但畢竟是業界第一份，而且是微軟下場參與認證的，所以我非常建議你可以學習和考一個備著。

注：因為領英目前已經退出了中國國內的運營，所有你會需要一個國際版的領英賬號才能參與學習和考試，如果你沒有的話，賬號註冊過程可能會有一定的賬號。

課程連結：https://www.linkedin.com/learning/paths/career-essentials-in-generative-ai-by-microsoft-and-linkedin

B站課程介紹影片：https://www.bilibili.com/video/BV1kh4y177Lu/?vd_source=b8ed956bbe6170c4bb6e401414e9d3ee

參考問題&答案​

在我影片發出的一個多星期內，我從評論區看到這個考試對普通人而言難度確實有點大，所以為了幫各位考過，我就提供一下我考試時的部分題目和我的答案供參考，實際每個人的考題構成和順序也許可能會不一樣。

  1. 問題：Your company produces science fiction and fantasy graphic novels. One of your top illustrators has developed a style that is very strongly associated with your brand. Your company decides to create a generative Al model to mimic their illustrations. This new model can create new graphics in their style in seconds. Now the company will have better control over their brand and increase productivity. What is one of the main challenges with this approach?

  • Current generative Al models are not doing a very good job mimicking creative illustrators.

  • It is currently illegal in the United States to mimic the style of working illustrators. The generative Al model will always need to be further trained, so it doesn't save any time.

  • It will "normalize mediocrity"-the graphics will look the same and lack a creative spark.

答案：It will "normalize mediocrity"-the graphics will look the same and lack a creative spark. ("這將'使平庸變得正常化' - 圖形將看起來相同，缺乏創造性的火花")

  1. 問題：You recently purchased a new smartwatch. To set up the watch, you had to go to the manufacturer's website and create a new account. When you create an account, it presents a long license agreement you have to accept to create an account. You were anxious to use your new watch, so you didn't scroll through the 50-plus pages of the license agreement. What is one of the ethical issues with how the smartwatch manufacturer is operating?

  • It should shorten the license agreement.

  • They are not sharing a clear and transparent privacy policy.

  • It should have you automatically accept an agreement when you purchase the watch.

  • It should always try to keep the private data on the watch itself.

答案：It should always try to keep the private data on the watch itself. (應該始終嘗試將私人資料保留在手錶本身上。)

  1. 問題：What is GAN？

  • when two generative Al organizations compete for the same resources

  • when two neural networks work cooperatively to produce the best output

  • when a discriminator generates output so that a generator can review it and offer adversarial feedback

  • when two neural networks work in opposition, with a generator and a discriminator to improve the generative output

答案：when two neural networks work in opposition, with a generator and a discriminator to improve the generative output (兩個神經網路對抗工作，透過生成器和判別器來提升生成輸出的質量)

  1. 問題：As a software developer, your Al ethical responsibility is to make sure that the Al nurse

  • is developed in a way that's transparent, explainable, and accountable

  • is well-versed in alternative forms of treatment

  • is always focused on generating data, increasing profits and reliable customer service

  • is always using the latest information

答案：is developed in a way that's transparent, explainable, and accountable (以透明、可解釋、可負責任的方式開發)

  1. 問題：What type of technology will your music player use to find your song?

  • It is using reinforcement learning to create a personalized playlist for your music player.

  • It is using unsupervised machine learning to make a recommendation based on your music tastes.

  • It is using generative Al to compose a new song.

  • It is converting your description into a search by using natural language processing.

答案：It is converting your description into a search by using natural language processing. (利用自然語言處理將你的描述轉化為搜尋)

  1. 問題：What's one of the key dangers for organizations that over rely on generative Al systems?

  • Generative Al systems will start to run these organizations with little to no human oversight.

  • They will regenerate the same material without any spark of creativity.

  • Your employees might resign if they feel that your system is in danger of replacing their livelihood.

  • Generative Al systems might make key decisions about who works for the company.

答案：They will regenerate the same material without any spark of creativity. (他們將重複生成同樣的素材，而沒有任何創新性)

  1. 問題：What machine learning algorithm are you using?

  • K-means clustering

  • K-nearest neighbor

  • Linear regression

  • Naive Bayes

答案：Naive Bayes (樸素貝葉斯)

  1. 問題：What type of machine learning are you using?

  • reinforcement learning unsupervised clustering

  • variational auto encoding generative Al

  • supervised learning multiclass classification

  • unsupervised learning binary classification

答案：supervised learning multiclass classification (監督學習多類分類)

  1. 問題：How is an artificial neural network related to machine learning?

  • an artificial neural network uses preprogrammed responses instead of learning.

  • an artificial neural network is an earlier form of machine learning.

  • an artificial neural network does not require programming like a machine learning system.

  • an artificial neural network is a machine learning technique.

答案：an artificial neural network is a machine learning technique. (人工神經網路是一種機器學習技術)

  1. 問題：What are the most essential ethical considerations to balance?

  • cost of implementing these new systems against the costs of maintaining full employment

  • getting creative generative Al output and optimizing production while maintaining human oversight

  • the dangers to humanity against the possibility of your own enrichment

  • your organization's obligation to appease shareholders against your obligations to humanity

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：What Al ethics violation might your chatbot release have caused?

  • Your technology assisted a human rights violation.

  • There is now a danger of competition from a large well-funded government.

  • Governments should always be able to use your technology for whatever reason they want.

  • You have not properly protected your technology from misuse.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：How does ChatGPT make sense of the world?

  • It observes the world, mimics behavior, and then produces similar output.

  • It makes sense of the world by reading data, processing that data, and creating clusters of similar information.

  • It draws on a vast dataset of preexisting text and generates output based on patterns it has observed.

  • It learns the world by observing humans and mimics their behavior.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：If you asked the AI "what is your favorite color?" and the AI responded with "I don't have feelings or preferences, so I don't have a favorite color," is the AI being honest?

  • The AI is not being honest because it does have feelings.

  • The AI is not being honest because it does have preferences.

  • The AI is being honest. It is incapable of having feelings or preferences.

  • The AI is being honest because it is mimicking human behavior.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：What does an AI use to determine which results to display?

  • It uses a combination of your search history and the keywords you type in.

  • It uses the generator and discriminator to find the most relevant results.

  • It uses an ensemble of models to determine the best results.

  • It uses a mathematical equation to calculate the best result.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：A national newspaper reporter is writing a story on generative Al. As part of the story, they chat for hours with a new online generative chatbot. A few hours into the conversation, the chatbot tries to convince the reporter to leave his partner. The chatbot company said they don't know why it gave these responses and will limit conversations to 30 minutes. What might be one of the biggest ethical challenges with this system?

  • The system offered personal advice too soon in the conversation.

  • Chatbots shouldn't offer marital advice.

  • There isn't enough transparency into how the chatbot is responding.

  • The system has access to tremendous amounts of data, so it can offer hard but truthful advice.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：What does the term model mean in generative Al?

  • A model is a set of algorithms that have been trained on a data set.

  • A model is a data set of ethical issues.

  • A model is a generative Al that trains another artificial intelligence on a dataset.

  • A model is Al mimicking human behavior.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：What is a good description of how a machine learning system operates?

  • A system achieves artificial general intelligence by collating responses from experts in every field.

  • An Al system learns in a way that is consistent with its preprogrammed responses.

  • A system "learns" by observing patterns in massive datasets.

  • An Al system generates content as opposed to just classifying existing data.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：Your online movie-streaming business wants to create an artificial neural network that can recommend new movies based on what customers have already seen. The team creates a series of XY diagrams of different film genres. Then it puts the film rating along the X-axis and the duration that people watch on the Y-axis. It then makes a recommendation based on how close movies are to each other on the chart. What type of machine learning algorithm is the team using?

  • Naive Bayes

  • Q learning

  • K-nearest neighbor

  • reinforcement learning

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：You are going to use machine learning to try and do a better job predicting the weather. To start out, you just want to classify two weather events: "rain" or "not rain." What steps would you take to build this system?

  • Use reinforcement learning to allow the machine to experience the weather first hand.

  • Build an unsupervised machine learning system that can cluster all weather events.

  • Train a supervised learning binary classification system with past weather data.

  • Use generative Al to create a whole new weather system.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：You are a technical manager for a large city courthouse. The judges have asked you to implement a new system that will make criminal sentencing recommendations. As part of your testing, your team has the system make sentencing recommendations for past court convictions. Your team finds that the new system is much more likely to recommend longer sentences for some groups of people. What is the main ethical challenge with implementing this system?

  • Impartial judges should make sentencing recommendations. Al systems should not be involved.

  • The courthouse obviously does not have the technical expertise to improve the system.

  • It magnifies existing biases rather than mitigating them.

  • The city courthouse might not be able to afford the service.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：Your company wants to create a smartphone application that identifies plants using the phone's camera. The company purchases millions of digital images of plants labeled with the species names. You use this initial batch of images to train your artificial neural network. What type of machine learning are you using for your network?

  • unsupervised machine learning

  • reinforcement learning

  • supervised learning

  • self-supervised learning

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：You are an executive for a large company that has a customer service department. Recently some of the top managers have been talking about replacing customer service representatives with an Al chatbot. Some of the managers feel like the chatbot should impersonate a human customer service representative. They argue that if customers know it's an Al chatbot then they would immediately disconnect. Other managers feel like it would be unethical to impersonate a human. What would be the best place to communicate your decision?

  • Create a Responsible Al Policy and Governance framework.

  • Schedule a company-wide meeting.

  • Create a Profitability with Generative Al Action Plan document.

  • The executives should leave it to the product development team.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：Typically, what are the three layers of an artificial neural network?

  • the supervised layer, the unsupervised layers, and the reinforcement layer

  • the input layer, many hidden layers, and the output layer

  • the transaction layer, the generator layer, and the final layer

  • the artificial layer, the machine learning layer, and the data layer

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：You're trying to get better at prompt engineering, so you decided to try a new technique. You say, "Write a 500-word essay on large language models and hallucinations from the perspective of a computer science graduate student at a university." What technique are you using here?

  • are taking an adversarial approach to get both sides of the story.

  • You are using role-playing to get more accurate responses.

  • You're brainstorming with the system about large language models and hallucinations.

  • You are using a compression technique by limiting the results to 500 words.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e

  1. 問題：Why did early artificial intelligence systems do so well with board games?

  • Because board games give the system unique insight into human behavior, early systems could learn and mimic the same behavior.

  • Because board games are inherently chaotic, the system had a lot of opportunities to crunch new data.

  • Because even with their limiting processing power, early systems thrived in a world of simple rules and pattern matching.

  • Because computer scientists could do a good job programming all the rules into the game that the system would understand.

更多答案可從我的專欄獲取：https://xiaobot.net/post/d3174f88-597a-47ca-91bc-a3e88e9a7f5e