This discussion is based off the PBS Frontline documentary, “In the Age of AI.” To watch this documentary, visit this link:
PBS Frontline: In the Age of AI
Questions for Discussion
- When Kai-Fu Lee says, “Data is the new oil,” and “China is the Saudi Arabia of data,” what does he mean in those statements? Can you describe at least three ways that companies or governments with data monopolies could benefit from having those resources over other companies or governments?
- From your perspective as a user or non-user of social media or search indexing technology, do you think social media, searching or website (cookies) tracking of US citizens is important in the race to become a leader in artificial intelligence for the United States, or is that tracking/collection an invasion of privacy and individual rights in our country? What about in other countries?
- Can you think of a few careers you might consider pursuing when you graduate from college or graduate school? Once you have a few ideas, visit https://willrobotstakemyjob.com/ and plug in those career ideas to see their overall risk to being replaced by automation. How vulnerable are those potential careers to automation? Can you think of 2-3 careers that existed in 1950 that do not exist today? How about 1990? How about 2020? List all the jobs from those years/decades that have become redundant and replaced by automation.
- We live in a remarkable time where nation-states are vying to become the leaders of AI technologies. What challenges do you foresee your generation will have to face when it comes to nation-states look to becoming AI leaders and potentially use that technology against other nations? Who do you see as the key players in that race? In your opinion, how will that shape diplomatic economic relationships between the United States, the European Union, the Russian federation states and China?
- How will services like ChatGPT, QuillBot, etc. change the landscape of learning a fundamental skill like how to write essays? What about writing code or undercutting the need to learn computer science? What other essential tangents of learning could be replaced by AI? Of those tangents, is it a good thing that humans would/would not need to learn that skill? Why? Explain your answers.
AS A REMINDER, please cite the URL of whatever sources you use to answer these questions.
1. One way companies could use data monopolies would be through targeted advertising which is already prevalent in the current day. Another could be price surging/changing based on calculated demand or on figuring out the maximum a customer would pay for a product. Lastly could be improving user experience and making for a better product which would help the product sell more.
2. Probably, but at the same time it’s also about privacy. The more data fed into neural networks, the better they turn out. But at the same time, there has to be more ethical ways to get data rather than invading the privacy of everyone.
3. I plugged in stuff such as developers and programmers and they had a pretty high chance of over 50-60%. There used to be milkmen deliveries that happened before refrigeration. There were also video store employees that stopped existing after newer forms of media came out. Recently, something like newspaper deliveries have probably been dying out. The popularity of digital forms of media including newspapers has all the more sent the physical newspaper business into a downward spiral. https://bestlifeonline.com/obsolete-jobs/
4. I see countries such as the US and China being big parts in this race. Both their infrastructures in technology are the most capable. In my opinion, I think countries that produce semiconductors and other useful technologies will either align themselves with the US or China and will cause a big war of AI.
5. I think that these services have changed the landscape of school forever. But just like the calculator, AI should be used as a tool rather than a cheat. It should assist. For example, in math class I still had to learn how to do multiplication manually. I think it is important that we learn the basics, but having a calculator allows me more complicated equations faster and thus helping me to do even more advanced problems.
Kai-Fu Lee describes data as the new oil because data is a valuable resource. Similar to how oil powers cars and electricity, data powers modern technologies such as AI and machine learning. He compares China to Saudi Arabia because, according to the documentary, China has a massive amount of data due to its large population and extensive collection of data. As the documentary notes, companies like Google greatly benefit from data because they can target customers and improve user experience. For these companies to successfully implement these technologies, they need data. But raises questions as to whether that data collection is an invasion of privacy. To some degree, I believe that it is necessary. It depends on what data they are collecting and how they are using it. Perhaps companies could be more transparent with how they use data, but that makes them susceptible to other competitors.
Some jobs that existed in 1950 but not today are switchboard operators and pinsetters for bowling alleys. An example of a job that existed in 1990 but not today is milk delivery. I could not find any jobs that existed in 2020 but not today, since automation typically costs money for businesses.
In the future, some challenges that will come with AI are increased competition between nations, greater technological disparities, and privacy concerns. The primary nations involved in the AI race will likely be the United States, China, EU, and Russia. The AI race could lead to increased competition and tensions between these nations. Alliances and rivalries could also potentially form, greatly impacting global relationships.
AI has already changed how we write essays. They can write essays, give suggestions, and offer feedback. But it’s not entirely clear to what extent AI can demonstrate critical thinking and analysis skills similar to humans. In my experience, AI can sometimes be helpful with writing code. It wasn’t particularly helpful for Android development because the code it generated was outdated or unconventional. This issue happens because dependencies and good practices are constantly being updated for Android development. However, it can be helpful with more fundamental concepts such as data structures and algorithms. Depending on the topic, AI could optimize research and quickly synthesize information. For now, I don’t think that humans will become obsolete in the next few years. As time goes on, we will continue to figure out how AI can be integrated into education and daily life.
Sources used:
https://willrobotstakemyjob.com/
https://studyworkgrow.com/20-jobs-that-dont-exist-anymore/
Kai-Fu Lee is comparing previous historical commodities of power—in this case oil and how ownership of oil impacted the countries that had it—to the growing importance of data in the modern world. Specifically, Saudi Arabia is a location with vast amounts of oil which has propelled them up the economic and international power food chains. Thus, Kai-Fu Lee’s comparison of China to Saudi Arabia predicts that there is a similar importance and power of the data that China has due to its high population and authoritarian regime’s focus on data collection. One such avenue which might highlight the importance of data are facial recognition and military spaces. With China’s burgeoning facial recognition sector, fueled by the data China has amassed, growing at a rapid pace, one can only imagine how the Chinese military might be able to leverage the technology to track targets through satellites, improve their spying on other countries, or anything of the sort. In a more economic sense, the data China controls allows it to train AI models much faster. As AI becomes more and more important to the valuations of companies, to the point where putting “AI” on a product vastly boosts a company’s share’s worth, China’s ability to improve Artificial Intelligence is destined to be a strength economically. Of course, it is not just China with this data advantage. Massive companies in the west also have a dramatic amount of data on the average consumer and this will benefit them as well. Advertising is becoming increasingly dependent upon data science to provide accurate insights on how marking is performing. This puts social media apps, such as Facebook (and their conglomeration of apps), Tik Tok, or even smaller apps such as Be Real in an incredible position to leverage their data to make large amounts of wealth.
I think the answer to both questions is yes. Yes, the tracking of data from US citizens is crucial to the United States’ efforts to maintain control on the international community; however, I also think that collection of data, especially if secretly and without permission, is an invasion of privacy and should not happen. This puts the US in a tricky position where we must choose to either lose international authority or break the ideas of our own constitution in order to keep up. If I am being honest, based on historical precedent, I would not be shocked if the US in the coming years (or probably already) decides to break individual rights in order to hold its power. I think other countries face the same issue: cause privacy concerns or be overtaken technologically.
Web development is at a medium risk of being made redundant by AI, while mechanical engineers are considered low risk; however, ultimately I am not too worried about AI taking my job opportunities. This is because I believe being a “strong thinker” (so to say) will help anyone land a job in whatever changing environment we live in. This really informs how I look at the colleges I apply to as the skills I learn may not inform my job just a couple years after I graduate. For example, human calculators which were used quite a bit during early spaceflight are all but non-existent today. Similarly, phone operators don’t exist anymore. These are only two examples of many, but ultimately the people who worked these jobs found other jobs for themselves. The market adjusted just as I believe it will after AI makes its impact.
As I kind of touched on in an earlier question, I think ultimately the world might be faced with the question of “is power or privacy more important.” I think the US, Russia, and China will all end up on the side of power, while the steps Europe generally has taken toward more privacy and less monopolistic tech companies makes me believe they might end up on the side of privacy.
I think it is a helpful metaphor to compare calculators to the new AI tools. When calculators were made more mainstream in a form factor people in school could access, people thought it was the end of the world as we knew it: People wouldn’t understand math anymore because everything was being done for them. This, of course, didn’t happen. We still learn the basics of math before we are allowed to use calculators, but then we are allowed to use the tools that help us speed up the tedious processes later in our schooling journey. I think everyone will still need to learn how to write (and they will), but soon enough teachers will be more comfortable allowing students to use AI to speed up the more repetitive parts of assignments. Ultimately we as students must learn how to live in a world with AI. It is not going to be cheating to use AI in the real world, so eventually we have to learn how to use it in class.
1. Kaifu Lee explains that having information itself is empowering to nations and multinational corporations, because information itself is all that is required to develop technologies. Information is the raw resource that can be processed into weapons, services, and energy. Just like how petroleum powers the world right now, and just as how theyhave for the past decades, it is information that corporations and nations will be competing for in the coming years. For example, China’s megvii facial recognition algorithm was shown to be more accurate than amazon and facebook. Part of the reason is China’s huge population, and hence their huge dataset their algorithm can train from. China is integrating these faciql reocngition to build a world without physical purchasing and manually enforced security. Another example is a company like amazon holding monopolies in automation of transportation and other replacement for blue collar jobs. Having information to train their programs and technology will in turn take control of the monopoly of the market further allowing them to develop even more advanced technology, putting them even more ahead in the race. Lastly, I came across an article about how rich reddit was getting as language models like chatgpt get popularized and developed. Reddit’s large data of human-generated articles, comments, and posts are authentic human data that language models and their developers want and need to train their algorithms. As a result, reddit was getting offered a lot of money just by having those human data with them.
https://www.nytimes.com/2023/04/18/technology/reddit-ai-openai-google.html
2. Clearly, there is a grey area surrounding this issue. It will definitely help the government to stay ahead in the arms race of AI, but there will always be a moral pushback of the people. It is best that the people trust the government to use the information for lawful and just purposes. Personally, I wouldn’t have a problem if there was such guarantee. However, politics and intervention of corporations would interfere with such a vision. Must the government start selling those information to companies who will use it for far more sinister purposes, thats when it would hit the fan. Another big example is China. China has seen quite a success with implementing these systems, but to an extent where they blatantly tread on citizen’s rights.
3. One of the things that I want to pursue is computational chemistry, as chemistry compared to biology or physics, has less computation integration. I always believed that the whole meme of “computer scientists being homeless” was overexaggerated. Nvidia’s ceo openly told the public to not study computer science anymore. Yet nvidia holds about 8 interviews for their software engineers. Google translate is getting better and prominent, but that hardly means people can stop learning foreign languages. Hence, chemical engineers/analysts are at low risk of getting replaced even though general consensus would claim that it would be replaced tomorrow. In the 50s, knockerups and “calculators” were prominent jobs. Knockerups made a significant amount of money to waqke people up in the morning. “Calculators” were a crucial part of the engineering industry where people did complex calculations at a time when calculators were not a thing. Today, our cell phone has replaced both of these professions, and arguable does it better than people. In the 90s, there was a profession in Korea called bus guide who would tell the passengers which station this was and where it connects to; now it is done by automation. My mom was a translator before she got married, and she still does some translator work as well. But she admits that translators will no longer be a thing in the next decade; AI is getting better at understanding and translating jokes and nuanced speech, in a far more quick and accurate way than human beings.
4. One bright side to this new trend, I suspect, is that there will be less of a risk of a physical nuclear major global conflict. The downside is that we may be nearing something a lot worse. People themselves would be more valuable, and primary skills such as penmanship and memory would become emphasized in the informational war. Overall, there will be a distrust of everything and everyone, and the destruction of traditional allyship of countries, since everyone would know that nothing is “open” anymore. Since everything is tracked and monitored, governments will hold more secrets to themselves without recording, resulting in an overall palpable tensions across the world. Aside from the US and China, I think nations with a lot of population or human interactions would have the edge in the race. This counts brazil, indonesia, or egypt. Also nations like swiss, norway, and netherlands who don’t have a lot of population but a lot of data and trade passing through them, would also be able to weaponize their unique strenght in the information war.
5. Just like my foreign language analogy, I don’t think the presence of AI will be able to convince the human race that writing or learning would be worthless. However, the focus of learning will shift from the content of what people to the way in which people learn to process information. Writing is not so much the content being written, but that act of recording itself, and the act of interacting with the thought to physical visualization. Developing the ability to interact and work with limited information and make conjectures would be the “point” of learning. Similar to this, just because computers can build our algorithms, doesn’t mean people don’t have to leanr those languages. If anything, the more widespread prominence of computer languages will induce people to learn more of them. They will make up more of the world, building the need for people to understand their world through them.
Like oil, the whole world will be increasingly reliant on AI, and China, from a combination of more people and more tracking, as more data. Data powers AI, and more data is more accuracy. Predictive accuracy is crucial to both companies and governments, and China’s models could be the best in the world just because of the quantity of data they have. Companies with data monopolies can use predictive modeling to target advertisements to users, tailor services to each specific user, and sell users’ data. Governments can can use it to track/predict/respond to crimes, target military recruitments, and predict opponents’ actions in war.
Personally, I have no issue with companies selling my data. I tend to reject cookies, but at the end of the day, I think predictive modeling is powerful and can improve our lives a lot. I do think there are a lot of unethical usages of data, but there’s unethical usages of anything powerful, so I don’t see it as particularly worse than anything else. I think in general, data has a lot of power to do good, and I think it’s okay for companies to track what I do on websites and social media. I do think there’s a line though. In a situation where loans are dependent on things as trivial as phone battery, I think it’s a step too far. I think things like government and banking should be separated from that level of data collection. I think of everything on my devices as personal life, so while I don’t see an invasion of privacy if computers look at it, I would like to keep it separate from my financial and legal life.
Statisticians: 54%
Musicians: 44%
Jobs that are at least partially automated: data filing/organizing (with physical files), mass mailing (replaced by emails or more automatic physical mail)
I think in some ways, countries will keep each other accountable. I know there’s a lot of ways people will use tech against each other, but I also think countries will recognize its power and keep each other in check. We’ve already seen this with current technology, and I think it will continue with AI. Could we all use chemical weapons on each other and create an apocalypse? Yes. Could we all use nuclear weapons on each other and create an apocalypse? Yes. But we have moral commitments to each other that have led to treaties and alliances, and I think that will continue with AI.
I think it’s hard to predict what new jobs will exist as a result of AI. There’s so many fields now that couldn’t exist before the calculator; we won’t really know until we keep practicing what we can do with AI. Certain things I don’t think AI will replace. ChatGPT can write essays, but it can’t write anything personal, and humans love reading personal things because they relate to them. AI-generated content comes from so many sources, it would be hard for it to have the level of personality that comes from one specific person having one specific life. The same goes for art. And computer science will definitely change. I think we’ll just be able to do more. There were mathematicians before calculators and there still are now. The field is still there; the people just do more advanced things because they don’t need to expend so much time and brain power on long division.
1. I think he means that one of the most needed resources in the near future will be data, as countries race to build AI they need mass amounts of data to build it off of. This is similar to how Saudi Arabia became rich due to innovation requiring more and more oil and them having excess oil. First, these countries can use the data to spread their influence as some countries will have systems reliant on other countries, like China trying to help other countries. Second, this will also help the countries that already have a lot of data as it will allow them to gather data from the other countries they are helping, growing this monopoly. Third, it would help with mass wealth accumulation and if there is ever a new cold war I believe the people on each side will be likely decided by who uses AI to become a new world superpower.
2. I think collecting data is a massive invasion of privacy, especially with the data you are unaware they are collecting. With the invention of new models that can predict so many things about you, I think this could definitely become dangerous. However, I also think this will be one of the most effective ways to collect data and improve AI. The tracking allows for much more data to be collected and thus allows AI progression at the cost of privacy. This question is definitely a basis of opinion, about whether you are willing to sacrifice a lot of privacy for technological progression.
3. Computer Programmers: 67%
A few I can think of is a lot of assembly line workers are being replaced with robot arms and self-checkout replacing a lot of cashiers.
4. I think the biggest threat from other countries AI is the potential for misinformation. As AI gets more and more accurate, it gets easier to create things that look real enough to fool most people, and quite frankly we are almost at that point right now. Likely the EU will be the first to implement laws against this, and I think that many others will follow suit soon after. I see the US and China becoming the leaders of it first as the EU has more regulations on data collecting where countries that have less laws about data collection will advance much faster.
5. Many of these skills will still be taught in schools, just not to the extent they are taught now. For example, we learn so many math skills that can be done by calculators as a way of learning how to think. With essays, we might not be writing them for quite as long, but it will still be taught as a skill to almost “train your brain”. I think that as long as people understand how to do it on their own, it would be fine to have some of these skills not be taught as long but I disagree with removing it entirely. I also think if the idea of humans monitoring AI writing code actually happens, we will still need to learn how to do it in order to detect mistakes.
What Kai-Fu Lee is saying is that data has a pivotal role in driving contemporary economic and technological advancements, much like oil did in the industrial era. Companies or governments with data monopolies can develop superior AI technologies by leveraging vast datasets to train more accurate models. This can allow them to gain consumer insights to personalize products and strategies, making them way more effective. Essentially they could have the ability to predict what you will do. And they can also make more informed strategic decisions to optimize their policies and investments. They can make more money.
The tracking and collection of data through social media, search engines, and website cookies play a pivotal role in advancing AI technologies. In the US, this data enables companies to develop more sophisticated AI models, which can have immense benefits. However, this method of data collection often causes significant concern about privacy and individual rights. The trade-off between leveraging data for technological progress and protecting personal privacy is a debated topic because it can erode trust and autonomy. In the European Union, for example, there are regulations such as the General Data Protection Regulation (GDPR) which are in place to safeguard personal data and enforce transparency. Putting emphasis on privacy rights over unchecked data collection.
I put in Anesthesiologists and it says that they are at low risk, 25%. Then I plugged in Models and it said that they are at high risk of being replaced by robots. After that I tried Actors and they are at low risk. Curious. 1950. Loom Fixer – Responsible for maintaining and repairing looms in textile mills. Automation and modern machinery have largely replaced these roles. Switchboard Operator – Managed manual telephone switchboards to connect calls. This job has been replaced by automated telephone systems and digital technology. Milkman – Delivered milk directly to homes. While some local deliveries still exist, this role has largely been replaced by supermarkets and refrigerated transport. 1990. Typist – Specialized in typing documents on typewriters or early computers. The role has largely been replaced by word processors and general office workers who use computers. Film Projectionist – Operated film projectors in movie theaters. Digital projection technology has largely replaced this job. Travel Agent – Provided personalized travel planning and booking services. While travel agents still exist, many tasks have shifted to online booking platforms and automated systems. 2020. Data Entry Clerk – Performed manual data input tasks. Automation and sophisticated software have significantly reduced the need for this role. Bank Teller – Handled in-person transactions and customer service at banks. Many of these tasks have been automated through online banking and ATMs. Retail Cashier – Managed checkout processes in stores. Self-checkout kiosks and online shopping have reduced the demand for traditional cashier roles.
The race for AI dominance can exacerbate geopolitical tensions, as nations may view AI advancements as critical to their strategic and military capabilities. Currently we have the fear of nuclear warfare, but I believe the potential for AI to be integrated into current warfare methods even more terrifying. Also different countries have varying standards for AI ethics and regulation. Discrepancies in how AI is governed and used can create friction between nations, especially if AI technologies are used in ways that contravene international norms or human rights. The proliferation of AI also increases vulnerabilities to cyberattacks and data breaches. Nations with advanced AI capabilities may face heightened risks of espionage and sabotage. In my opinion, key players in the AI race include the United States, China, and the European Union. The U.S. is known for its leading tech companies and innovation ecosystem. China has aggressively invested in AI as part of its strategic goals. The EU emphasizes ethical standards and regulatory frameworks. This competition and difference in viewpoints could cause further conflict amongst these countries.
Generative AI services have and will drastically change our learning systems. WHile there is an effort to limit its access in school systems, the use of it is kind of inevitable. Relying heavily on AI might lead to a superficial understanding of subjects, as users might not engage deeply with the material. And overly relying on AI tools can erode the development of critical thinking and problem-solving skills, which are essential for comprehensive education and professional success. While generative AI is very bad for learning, developing, and maintaining critical skills and increases our reliance on technology, it has its benefits and positives as well. It can, sometimes, be more effective than a human. Like the example in the documentary, using ML models can save so many women from breast cancer. Also The ability to type caused a shift from handwriting notes and essays. How many people today know how to write in cursive? These kinds of shifts have happened before but no one will say that we are not benefited by it. Moving away from paper and onto technology helps out with sustainability, which wasn’t as much of an issue back when. So as times change, and the need for certain things rise or go away, the way we work and learn with technology will change with it.
Sources:
https://willrobotstakemyjob.com/
https://www.pbs.org/wgbh/frontline/documentary/in-the-age-of-ai/
He means that data is the new oil in that it is the new commodity that companies are looking for to train AI systems. He describes China as the new Saudi Arabia because they are known for having a lot of oil, and Kai-Fu Lee believes China has the most data because of a combination of development and total population, which leads it to have the largest amount of data. Three such ways companies can benefit is by creating more accurate representations of where they should build a restaurant based on search results and ordered food as China has 10x more ordered food than the US. Another way is for making loans, if a company had a data monopoly on people they would better be able to decipher who would be good investments and who wouldn’t be based on online activity and shopping habits. Similarly, if a company had a data monopoly on movement through location and passenger numbers per train, they would be able to better predict where to locate their next rain better than their competitors.
While I am ok with a lot of tracking I think there is a limit at which it should be and I think that this limit should be similar in and out of the country unless it is clear that the individual poses a clear threat to society. My main opinion is that they should not store sensitive information about you, but I would be fine with them tracking a lot of things including location.
While some of the more straightforward jobs such as computer programmer have a high risk, I found that jobs more about creativity and thinking and problem-solving have lower risk such as computer systems engineer or mathematician. One of the largest jobs that comes to mind is the days of the switchboard when people manually plugged through calls, which were replaced by dialing numbers. Similarly, factory jobs mainly in automobiles have taken a huge hit due to automation of manufacturing. In the 1990s one of the larges professions that went away was the mining industry as it was taken over more by machines than people. While I cannot think of any specific job that has taken a hit since COVID-19, I think a lot of the service industry has suffered from it though I don’t believe the dip is because of automation.
I think the largest three will be the combined EU led by the Netherlands and ASML, combined with the US in fighting China over AI. While I believe that there will be some competition, I doubt it will be as large as the Cold War, and instead will be competition over chip manufacturing capabilities though the large chip demands may be brought on by AI. While I think that Russia may play a role, I think their role will be much smaller than the others since they don’t have the same capacity, and they will instead focus more on cyber security and conventional military than AI.
I think and hope that the current education system turns into a system of thinking, and learning how to think and problem-solve. I believe this because these AI sources will be able to do all of the smaller, more minute details, but will still need the bigger picture ideas, which rely on problem-solving capabilities more than knowledge of the area. I am personally a huge proponent of this type of learning because it allows everyone to contribute and be able to succeed, and will also limit the amount of useless memorization. I also think this is the kind of revolution that the educational system has needed and AI is the perfect opportunity to make this change as it makes some of the information unimportant.
Sources:
https://www.ibrc.indiana.edu/ibr/2002/fall02/fall02_art4.html
https://willrobotstakemyjob.com/
Saudi Arabia was the main producer around the world for oil, and here Kai-Fu Lee means that China is now the one who is in control of data and its distribution to other countries in the world. Companies that have data monopolies can buy better ads to reach potential customers better as the data would allow them to know who would be potentially interested in their products. Governments with data monopolies would also know their population demographic better. This means that when they push a political agenda, they know the best way to distribute this information like abasing other government images and showing their relatives/friends’ beliefs. The same could apply to companies and not just their advertisement but also the products that they make. They are aware of what their customers want out of them and what improvements to make in order to make their product a better competitor in the market. Companies without this data might not make the most wanted products or the products with the upgrade the customers liked the most. Therefore, having this information allows companies to be the best competitors in their market,
I think that when the company makes clear what information they’re tracking, what cookies, when, and how they do so and allows the user to give consent to the cookies then it is not an invasion of privacy because the user is aware of what personal information is being recorded. I don’t think social media, searching, or website tracking is absolutely vital to becoming a leader in artificial intelligence. Artificial intelligence doesn’t always mean that we are using machine learning to push advertisements to keep security checks on citizens. Artificial intelligence can be used in different ways like data sorting and playing video/board games like Go. These other formats don’t require tracking of personal information in the country and can be optimized without social media or search indexing technology. Our country’s focus is to use artificial intelligence to help the people, but there are ways of voluntary tracking or ways of tracking that are more open about how one’s information is being used which I feel is a better way of allowing AI to advance in the country. For example, self-driving technology for trucks doesn’t require knowing what advertisements I like to click on or what products I’m interested in, so I feel like social media and related stuff aren’t required for the US to advance. However, because our focus of artificial intelligence is to help improve people’s lives then I do think that it would be necessary to have the private information of citizens recorded. I think the government does have to be completely open about what information is documented and how it would be used and allow the people to have a choice in saying whether or not they want their information to be used. I think if citizens don’t even want their information being used, then there’s a small chance they would want their information being used by the AI in the future. I think the same applies to other countries.
The ideas I might consider pursuing on the website are Radiologist which has a 39% of being replaced, Chemist: 41%, and Pharmaceutical scientist: 47%. By the 1950s, I think lamp lighters and car makers for carriages would have decreased tremendously in popularity. By 1990, typists and telephone operators would have significantly decreased in job availability. I don’t think there are any jobs that exist in 2020 that don’t anymore, but there are jobs decreasing in demand like people working in machine assembling.
I think the challenges our generation will face are with the decrease in privacy of the increase in AI technologies AI requires data and often that data is of our private information that people might not want to be used. I think there is going to be a conflict between people wanting to move on and improve AI and others who don’t want their personal information used. I think people would want to get ahead of other nations but then others wouldn’t want to expose/release their own information. I think key players in the race are going to be the government and investors because AI technology will not go very far without any investments. I think it would cause economic relationships to strain and as well as lean towards countries/nations with more AI technology as countries want to be able to advance their countries further. Since more and more nations find AI as the future in being able to help one’s country, they’re going to align themselves with the nation that is at the head of that. Therefore, countries that are already technologically advanced would have increased tension as they are fighting to be the leader in AI.
I think AI services can help students in the beginning on how to write according to a structure and it can also provide, though I’m not sure of the accuracy, feedback depending on one’s essays (giving tips on the strengths and weaknesses of a student’s essay). However, for creativity and learning how to write with one’s own voice, I believe AI services can not help very much because that is something that the writer must develop for themselves and AI can’t write an essay for them. I think these services can help students develop a skill, but these services would also provide lazier students with a way to cheat and just have an AI write the essay for them. I think that learning to write code is still important because it is a way of problem-solving and strengthening mental skills. Just because we have calculators now doesn’t mean that we didn’t teach students how to do addition and subtraction when they were younger. AI being able to replace a skill doesn’t necessarily mean humans “need” or “not need” to learn a skill.
resources:
https://www.midwestteachersinstitute.org/ai-will-transform-teaching-and-learning-lets-get-it-right/
Data is the “fuel” for improving generative AI models, so the more data that you can feed your AI, the smarter and more efficient it will become. Similarly, oil is literally the fuel for thousands of products, such as gasoline, plastics, asphalt, and energy. Saudi Arabia is a huge exporter of oil with their vast oil reserves, and China’s huge population and ability to collect data on every citizen gives it access to vast amounts of data they can feed AI with. This new edge in information and resources will propel China to more international power and rival other superpowers with its leverage. Many companies, especially social media platforms, have access to similar amounts of data through their user base. Social media platforms can track thousands of different habits or actions users perform, and then use that data to improve their services, market certain products, or even push certain agendas. Companies with more data will also be able to generate more money through tactics like targeted advertising, where certain users will be marketed certain products. Lastly, as AI is becoming more integrated throughout many sectors, data will be greatly valued by companies and governments with its ability to train generative AI models and hasten progress.
I would say that there needs to be a certain balance between the two, and we shouldn’t have to pick between one extreme and the other. When I sign up for social media or use the web, I also acknowledge that my data will be tracked and analyzed by either a private company or the government. At the same time, my privacy is essential to how comfortable I am using online services, and it does make me a little uncomfortable, even if my data may help the US gain an edge in the artificial intelligence sector. What’s most important is what my data is being used for, and whoever is tracking my data must be transparent about it. For example, if Instagram is tracking my data to target ads, improve their platform, and learn user trends, then I’m okay with that. In fact, I prefer getting ads that I might actually care about. On the other hand, if Instagram were to be selling my private information without my knowledge, I would consider that an invasion of my privacy. One person in the documentary claimed that “nerds say they wished they came up with that graph”, and I’m not sure how accurate that is, but the more data being collected and the more types of data being collected, then more graphs can be graphed and more trends can be found. Then certainly social media and website tracking will help the US in the artificial intelligence industry. In other countries, there may be a different culture around government tracking. For example, I assume South Korea would be more open to tracking because their used to it through their vast CCTV system.
One of my dream jobs is becoming a fast food worker. Unfortunately, fast food workers are at a 92% risk of being taken over by robots. This is likely due to new technology such as online ordering, thus eliminating the need for hosts and waiters. At some restaurants, there are even robots that serve food, and I wouldn’t be surprised if there are new developments where robots can cook relatively easy meals at fast food restaurants. Fast food workers may find themselves without a job in the next few decades. In 1950, there were professions like newspaper boys, telephone operators, and assembly line workers that do not or hardly exist today. Newspaper boys don’t exist because of online news and the lack of newspapers. Telephone operators don’t exist because anybody with a cell phone can call, and assembly line workers were replaced by more efficient workers. In 1990, there were camera film developers who would develop the film on old cameras until digital cameras were introduced in the 1990s. In 2020, travel agents have become a lot less prevalent, partly thanks to Covid and mostly thanks to online services that allow you to book trips without an agent.
Since data is so important in the development of AI, I foresee our generation having to deal with less privacy in exchange for data collection done by governments. That may be difficult in a country like the United States where freedom and personal rights are so highly regarded and valued versus a country like China where monitoring of the citizens is commonplace. The experts in the video considered AI as a huge revolution that will change the world greatly and rapidly. On the extreme side of things, it’s possible that many world superpowers will enter a race to the strongest AI, similar to the race to nuclear weapons and power. On the more extreme side of things, there could be another Cold War except for AI. As far as the countries that are key players, countries that are major exporters of semiconductors will benefit greatly from exports, the top four being China, Japan, Malaysia, and Germany. The United States is also very technologically advanced, and will likely hold a big role in this race to AI technology.
Generative AI like ChatGPT have been used by many students to do their work for them, ruining opportunities for students to use critical thinking skills themselves. If generative AI becomes so good to the point it can effectively simulate a person’s writing style without being detected, then there may need to be a shift to more creative assignments that involve things other than just words on a page. Learning how to code may be very important in the future as technology and AI become integrated in almost every field and profession, and understanding how AI and code works will be very important. On the other hand, if AI can start to generate basic code by itself, then that will undermine the need to learn how to do basic code. I think this can either help students dive in to more advanced topics immediately, or ruin the gateway for computer scientists to learn basic code.
Kai Fu Lee means that China has a lot of data, and that data is the fuel of AI and the future. Companies and governments with data monopolies can benefit from these monopolies by making a lot of money. One way they can do this is to sell user’s data to other companies or people for money. Another way is to use this data for targeted adveratising, which is connected to the previous way. Finally, governments can create facial recognition and other algorithms when collecting data on people.
I don’t use social media for a number of reasons, but one of them is to avoid having data collected on me. I think that the tracking of US citizens is not important to become a leader in AI unless we want to base our society off of AI. It is an invasion of privacy to unknowingly have data collected on you, and people use misinformation to target adveratising and searches for you. This doesn’t really help society, but pushes consumerism.
Calculators, elevator operators, and switchboard operators aren’t jobs anymore. Also now the service industry is taking hits. United lounges have robots that collect dishes and give them to people in a systematic way. The only engineering career that I can think of that doesn’t have a risk of being replaced is mechanical. Even computer programmers are at risk of being replaced. System engineers have less of a chance of being replaced, but are still at a decent risk.
I see China and the US as major players in AI, however, as we’re seeing in the Russia-Ukraine war, any country can come up with innovative ways to use new technology. I think that our generation will have to face quite a few problems with AI becoming more popular, for example, AI based systems can have bias. Take tay.ai or mask recognition algorithms. The mask recognition had bias against african americans. I think that the economic relationships between the US and EU will stay roughly the same, but the US, Russia, and China relationship will deteriorate.
In my opinion, I think that if AI can replace every aspect of human society, we won’t have anything to do. If it writes software and can do a lot of our jobs, it will result in major job loss and politicians will just turn it into another issue to debate about. I think that learning these skills and using them is good for humanity, because we need to expand our minds in order to be better and moral people. If AI takes learning about these different skills away from us, humanity will be pointless
Kai-Fu Lee’s statement means that data is becoming a commodity that is both valuable and mass produced as Saudi Arabia drilling oil. China is at the forefront of gaining wealth from data, with advanced technology and data-oriented implements installed both by the government and private corporations that benefit the people. Companies or governments with monopolies on important datasets can better sell products by recommending more relevant products to consumers, make more informed and impactful budgeting decisions for both companies and governments, and also be able to better maintain social order and security in the case of governments having data. In the case of social order and security, China excels in using technology to create a safe public society yet respecting the private activities of its citizens.
Social media, web search and website cookies are honestly the very tip of the iceberg in relevant user information needed to create a smarter technological experience. For example, the digital instant loans provided by WEIXIN are only possible because of such a comprehensive collection of data, including device usage habits and pattern, contact info and text content. The US could never really catch up to China in the sense that most US citizens have this irrational sense of what absolutely needs to be private, for example refusing to let their web searches to even enter an anonymized database. Without the voluntary cooperation of technology users who recognize that data collection can only benefit them, as good standing citizens, artificial intelligence would only be a conflict between the people and companies like in the US, rather than a tool for and by the people.
In 1950, it was unheard of to earn money over the internet. Now, we have people whose sole jobs are to maintain websites, post blogs, and act as internet journalists and niche-reporters (see “CNX software”), making full-time salaries.
I pursue a college where I can specialize in electrical engineering, and the chance provided by the calculator is around 30%. However, I don’t see AI agents replacing subcategories such as RF/broadband communications system engineers (what I’m interested in), where extremely specialized knowledge and industry experience is both required, and not really “learnable” online.
The “AI-race” is really a pessimistic way to frame the global development of technologies that comes from a US nationalistic perspective (an attempt to frame leaders such as the PRC with different ideologies as enemies such as in the Cold War, echoes directly the space race). As for challenges in my generation as different organizations develop AI, I see the greatest challenge being navigating AI use in the workplace. More progressive workplaces encourage and even require AI use, sometimes resulting in poor standard designs or complicating the code toolchain (see “AIDEVOPS”). However, there are some workplaces imposing a blanket ban on AI GPT, where tasks such as collating spreadsheets can’t be done by AI and must rely on human skill.
I see a good future in terms of diplomatic relationships and collaboration between the Chinese government and the US. The US and Euro stance on providing China with the tools for training AI (see “NVIDIA 4090D”) is one currently fueled by blind hate and fear of China. I see in the near-middle term future that more and more China software and model technology will be exported for use in the States, and eventually Chinese AI development will lead by so much that sanctioning Chinese software won’t be a viable choice anymore for the US. This is similar to my country’s manufacturing ability in consumer goods and class-2 electronics, rising from farm villages to world leading manufacturing hubs.
I would never use ChatGPT to write a graded essay. Its writing style – although grammatically perfect – is bland and predictable. I believe good purposes of AI GPT for academics are note/audio transcript and summary, and being used as a smart database search tool. AI GPT as it develops in the future, will separate those who truly learn and those who use it as a crutch. Eventually in higher fields, this will show in those who chose to learn irresponsibly in their blind reliance on AI outputs rather than the use as a tool for one’s own creation.
Please give a full description of the nature of your first AI/ML game project.
My AI game project was a bot to play the game Hnefatafl, a viking board game sometimes nicknamed “viking chess”. This project was a collaboration with Ryan, hoping to have our bots play each other.
What was the steepest part of the learning curve for this project? Was it learning how to implement the AI or how to use the Pygame/Pyglet library? Please elaborate and explain your answer.
The hardest part for me was figuring out how to do the training along with PyGame. My solution was to not actually use the PyGame UI in my training, but that made it difficult to follow what was going on. I also didn’t have experience with neural networks before doing this project, and I had trouble conceptually understanding what goes on in a neural network and what type of network would work the best for my project. Because of the complicated spatial relationships in Hnefatafl, I chose to do Deep Q learning with a Convolutional Neural Network.
What went “right” with your project? As in, what worked seamlessly? What went “wrong” with your project? As in, what were your biggest hurdles or where did you have the most trouble debugging or getting your project to run?
I eventually got a bot that plays very well and beats me practically every time, but that was after a lot of model changing. There were several challenges though. One was that the two teams in Hnefatafl have different objectives, so I had to train separate models. Additionally, since Hnefatafl is a pretty obscure game, there’s no data to train on, so I had to train on self-play. The issue with self-play with my own models playing each other is that it’s hard to track progress of the model. Is one model doing better because it’s getting better or because the other one is getting worse? I did track win percentages over episodes and tried to make sure I didn’t have a model that was winning super rarely (<10% or so). I considered that a sign that one of my models was missing strategic moves, so I adjusted rewards and eventually decided to add simulations (explained in the next question).
Describe the AI/ML algorithm your game implements. Did you work through a tutorial you found online? Did you start from scratch because you were motivated by a particular game or algorithm and you wanted to implement it using Pygame/Pyglet?
I started from scratch and had some conversations with ChatGPT about approaches that might work for Hnefatafl. I started with a pretty simple DQN, but eventually realized that using a CNN in my DQN would help it understand Hnefatafl better. Later, I realized with the number of possible actions in Hnefatafal, I would have to use some simulation in addition to my reinforcement learning to get a model that made good moves. I designed my own simulation algorithm that creates a tree of possible future moves and then feeds its data into my DQN model’s decision making.
If you had to teach this class next year, what project would you recommend to students in the Advanced Topics class to give them a broad and comprehensive overview of some fundamental AI algorithms to implement in a game?
To be honest, I don’t really think doing a board game was the best implementation of reinforcement learning. I definitely think I got to learn about reinforcement learning, but I think I might have been able to get more out of reinforcement learning techniques with more of a video game than a board game.
Include your Github repo URL so your classmates can look at your code.
https://github.com/Ryan-Bauroth/Project02_PygAIme.git (see mooreo directory)
My project uses deep learning to optimize the way that it plays Othello, also called Revesi by some. Using epsilon to introduce randomness, the model transitions from exploration to exploitation. It uses a normal torch model as well.
I think it was generally the AI and trying to figure out whether the AI was good, since I was using a two-player game, it was hard to tell if they were getting better. As such, I found the hardest part to be evaluating my model, but also trying to understand some of the concepts such as the loss and how to implement it.
It worked well in the start because it was getting better, but latter on it got harder and harder to tell if it was getting better since I was only basing it off itself instead of trying to rain it against something like a minimax. Most of the bugs I ran into had to deal with the shape of my matrices because of the way I was inputting the data as a group of 30 8 x 8 boards so they sometimes got compressed or decompressed into 240 x 8 instead of 30 x 64.
I just browsed online and used the different sources to learn how to use deep Q learning. The reason I chose it was because it was the one that was recommended to start with because it was simple. Accordingly, I used it, though in the future I might want to try it with different models including something like monte carlo which Ryan used so it won’t be based entirely based on itself.
I would recommend a non-simple single-player, this is because it is easier to evaluate and also will allow them to experiment with different kinds of models. An example of this would be a jumping game similar to that of Aarav’s because it allows the person to try to maximize different aspects.
https://github.com/CameronJMorris/Project02_PygAIme.git