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 did he mean with those statements? Can you describe at least three ways that companies or governments with data monopolies have benefitted from having those resources over other companies or governments, especially now in 2026? Keeping in mind this documentary was released in 2019, most of it filmed from 2015 to 2018, how has data collection, mining, and prediction changed our world in both physical and virtual ways?
- 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? Should our government and private entities be more up front about data collection in the US (social media, Flock cameras, etc.)? What about in other countries? Why is data collection in the United States different than in more authoritarian states like China, etc.?
- 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, private companies, and publicly traded companies 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 have services like ChatGPT, Claude, Gemini, etc. changed the landscape of learning? You are currently enrolled in a computer science course. What made you want to learn computer science? What other essential foundations of learning do you think will be replaced by AI (if they haven’t already)? Of those foundations, do you think it is a good thing that humans would/would not need to learn that skill? Explain your answers.
AS A REMINDER, please cite the URL of whatever sources you use to answer these questions.
1. When Kai-Fu Lee says “Data is the new oil” and “China is the Saudi Arabia of data”, he is referring to the fact that in the future/present, data is used to train LLMs, and other AI models will become increasingly valuable, as valuable as oil is today. Because China is an authoritarian regime, it can collect data at an unprecedented rate compared to other countries, allowing it to build more powerful AI models. Some ways that companies with data monopolies can benefit are building stronger models more quickly than other nations, always having data that can retrain models that make mistakes (so an infinite improvement source), and, in the situation concerning larger and smaller companies, the larger companies force the smaller ones to use their data platforms to be able to keep pace with their competition. Data collection, mining, and prediction have drastically changed since 2019. Now, it’s commonplace for companies to collect your data when you visit their website, and this has led to mass amounts of political micro-targeting. Political micro-targeting is when campaigns are tailored to each individual’s preferences. For example, if someone has shown they care about the environment, the party might show headlines that a certain candidate is pro-environment. This has become very popular over time, and it’s how politicians maintain different facades for different people.
2. Social media or search engines tracking US citizens can be enforced to a certain degree, but due to our Constitution and Bill of Rights, tracking must be done constitutionally without invasions of privacy. Cookies are a decent way to do this; the user consents to their data being collected, which isn’t an invasion of privacy and also helps the US build its databases. AI will be a major part of the future, so the US needs to keep collecting datasets to improve its AI capabilities. Whether there remains a more fruitful avenue than civilian tracking remains to be seen. The US should be upfront about its tracking; otherwise, it could become a scandal like the NSA. US citizens should be kept in the loop about data collection, and being upfront about tracking is a good way to start that relationship. However, in an authoritarian state, data can be collected at will and without consequence because the party in power is absolute. This allows them to collect data on anyone whenever they want.
3. I want to work on Wall Street or as a quantitative analyst, but those types of jobs are heavily under threat from software and agents who can essentially do the same thing but more efficiently. I want to be adaptable and improve my human soft skills, since AI can’t build relationships with people well. According to the website, math-heavy jobs such as actuaries and quants are under slight risk, while investment bankers are under heavy threat. In 1950, 2-3 jobs that existed then no longer exist today: elevator operators, soda jerks, and telephone switchboard operators. In 1990 and 2000, some jobs that existed then no longer exist today are video rental companies, photo developers, and typesetters.
4. I think there’s a lot of political polarization in today’s era, and with the looming threat of AI, the political climate could prove to be a challenge. This could prove to be dangerous and threaten the lives of millions of people. I think the geopolitical big three (China, Russia, USA) will be major players in the AI race, as they are everywhere, and I think India, since it has so many people and a semi-authoritarian regime, will grow to be a major player as well. I think this will severely affect economic relationships between the major powers because any wrong move could have major consequences; it could become a game of chicken where mutually assured destruction goes away.
5. LLMs such as ChatGPT, Claude, and Gemini have changed the landscape of learning by changing student behavior and also the classroom learning environment. Students now rely on AI models to complete work and help them learn, and teachers are starting to integrate AI into their courses without simply letting students do whatever they want. I think CS is a great skill to have, and it’s challenging, which made me want to try it. I also think Machine Learning, vibe coding, etc will be very useful skills in the future even as AI levels up at coding. Just as calculators made learning arithmetic and multiplication close to redundant, I think AI will make lots of classes focus less on problem solving and more on theory. I think humans would still have to learn it, but the emphasis would be placed on other topics that are more valuable to critical thinking.
I’m interested in what you mean by classes focusing “less on problem solving and more on theory.” Isn’t the most important part of problem solving the theory behind it, i.e. the how, not the outcome of one particular problem? I think the problem with using AI to do what we might consider busy work is that that busy work is necessary to practice the skills to even consider theory in the first place. I think something like calculus makes a lot less sense if you can’t already quickly picture graphs in your mind, but that’s something you only learn to do by graphing a lot of functions by hand over and over (literally muscle memory) rather than using Desmos or the Google AI overview. Of course, this may not be unique to AI models—and I agree that teachers are integrating AI more to minimize ‘rogue’ usage—but I think we train critical thinking with our current curriculum for a reason, and AI threatens that.
Your point about India becoming a major player is one I hadn’t considered, since most people just default to the US and China. Do you think its advantage would come from population and data volume, or more from the talent pipeline, given how many engineers it exports? I’m curious whether raw data still matters that much now that the bottleneck seems to be compute and chips rather than training material.
I think your second point is interesting regarding the constitution and bill of rights. I think the US is passing some of those thresholds which could be considered as invasive but nothing is stopping that sort of data collection (eg Flock cameras) because these documents are so old that there’s a lot of leeway in how they are interpreted. It’ll be interesting to see different viewpoints on this issue going forth.
1. Data is now an important resource that can help make countries really powerful. Kai-Fu Lee is saying China has access to large amounts of data compared to other countries. Companies with data monopolies have the ability to make better AI models and can in turn outcompete any competitors’ models. Governments with data monopolies have a lot more information, and thus have the ability to better detect and prevent fraud and cybersecurity (https://www.sas.com/en_gb/insights/articles/big-data/big-data-government.html). Data collection and mining has become a lot more efficient since 2019, and there are significantly more data centers now. Virtually, platforms have the ability to give users very targeted and personalized content and ads now, and AI models have developed a lot since this documentary was made.
2. According to some minimal research, data collection through social media and the internet in general is very useful towards advancing artificial intelligence because it gives AI models access to very large datasets that come from real people. I think it’s a tad invasive; most people don’t realize the extent of data collection that is happening to them (both what kind of data and how much of it is being collected), so transparency would definitely help with this. Data collection is different in the US compared to places like China because in the US, data management is mainly carried out by private companies (and there’s also more of an emphasis on freedom of speech and privacy), while China “views data as a strategic resource requiring state control”, so its collection and use is mainly managed by the government (https://gradiant.org/en/blog/sovereign-data-spaces/).
3. I don’t really know what jobs I want to do, so I just put stuff that I was curious about: Animator (43%), Video game designer (22%), Lawyer (12%), Chef (29%)
Defunct jobs:
Switchboard operator, milkman, typist (1950) https://historyfacts.com/us-history/article/popular-1950s-jobs-now-obsolete/
Bowling pin setter, Blockbuster, encyclopedia salesman (1990)
Telemarketers (mostly), watch repair, bank tellers (2020)
4. I assume there’d be challenges with more invasive data collection. Key players in an AI race would likely mainly be the US and China. I have no idea how this would affect diplomatic economic relationships honestly but I guess it could strain the US-China relationship more and potentially develop collaboration between the US and the EU and between Russia and China.
5.
– ChatGPT, Claude, Gemini have made it really easy to synthesize information and speed up brainstorming/ideation, and students increasingly use these LLMs to learn
– I think computer science is pretty cool but also my parents wanted me to learn it
– Note-taking
– I think it’s important to know how to take notes just for the purpose of understanding what methods help you learn the best, but other than that I’m not sure it’s necessary
– Writing essays
– I think it’s probably better that humans should learn this skill – writing is how we can express ourselves and communicate information, and that can’t really just go away
I agree; beyond virtual platforms, tailoring each user’s experience to their interests has helped politicians and world leaders gain an advantage and create a wall of propaganda. Even in democratic countries like the US, politicians have used micro-targeting to tailor their campaigns to each individual person. In more authoritarian countries, propaganda isn’t as necessary since there’s no political opposition; authoritarian countries use propaganda mainly to stay in power, while in democracies, propaganda is mainly used to gain power but is also used to maintain it.
I also think a level of transparency would help the American public, since many people don’t know how much their data is tracked by major corporations.
I’m surprised that the legal profession had such a low % of being taken over by AI; lower-level associates whose main jobs are grunt work like spell-check and doc review can be replaced easily with software, while more senior management/lawyers will probably have a more difficult time being replaced by AI. 12% seems very low.
I’m curious about whether middle powers like the EU will deepen or break their relationship with the U.S. I agree that AI has the potential to threaten peaceful ties between the U.S. and China, but I think it’s also possible that countries that have historically been subordinate to the U.S. (think countries in Europe, Canada, South Korea, Japan, etc.) may also see this AI race—and the Trump administration’s attack on science—as an opportunity to elevate themselves to the level of major power. This could create a multipolar world where there are more than two countries with major stakes in the AI race, and each would be a threat to the other.
Your distinction between private data collection in the US and state-controlled data in China is a good one, but I’m not sure how much it holds up in practice, since the government can just buy data from private brokers instead of collecting it itself. Do you think that gap actually matters, or is it more of a procedural difference than a real limit on what the government can get to?
1. Kai-Fu Lee’s comparison works because oil was the resource that fueled the 20th-century’s economic and military power. Whoever controlled it could fuel industry, transportation, and war (oil was actually the reason Japan and the US fought during World II, Pearl Harbor was carried out as a reaction to American oil sanctions). Data plays that same role for AI because machine learning models only get better with more examples to learn from, so a company or nation sitting on vast amounts of data can train more accurate, more capable systems than ones that don’t. China fits the “Saudi Arabia” title because it has 1.4 billion people generating data constantly, and more loose privacy protections unlike in the democratic world which means the government and their affiliated companies can collect and use almost all of it without much legal friction. Three ways data monopolies have paid off: The United States has used its dominance in internet infrastructure, and venture capital to make sure the world’s leading AI labs such as OpenAI or Anthropic are American companies. Because so much of the internet’s text, video, and code is in English and hosted on US-based platforms, American companies had a head start on training data, and that head start compounded with American companies having better models, attracting more users, which generates more data, which trains even better models. That gives the US outsized influence over what AI globally looks like and what cultural values it reflects. Governments with access to massive datasets have used them for social control at a scale that wasn’t possible before. China’s use of facial recognition and data collection to run its social credit system, or to manage COVID-era contact tracing, lets the Chinese state model and predict citizen behavior in ways that governments that don’t surveil their own people can’t match. A lot has changed since they filmed it. Companies collect way more than they used to, and from more places. In 2018 it was mostly phones and browsers with a significantly lower percent of the global population being fully integrated into the digital world. On top of that, the models that turn all the data into predictions got much better in a relatively short time. The things in the documentary look honestly kind of basic next to what came after it, since almost all the big jumps in language models happened after the film came out.
2. I do think that data matters for the AI race because models learn from what people produce, so the US having huge platforms that billions of people use is part of why American labs got ahead in the first place. If that were all cut off, it would probably slow the country down. But I don’t think that’s a good enough reason to be like China. The whole point of the US is supposed to be that you get privacy and freedom by default, and the government has to justify itself to you instead of the other way around. If we win the AI race by surveilling our own citizens the way China does, I don’t think it’s worth it. To this end, the government and private companies should be way more upfront about what they collect. Flock cameras, for example, are problematic because people in a town with those cameras have no idea their license plate is being logged and stored every time they drive past one. It got installed without a vote and now there’s a searchable record of where everyone drives. A key difference is that China does data collection in the open and the state runs it directly. Chinese people know that cameras are almost always watching and the data feeds things like a social credit system which can cut off your access to regular daily life things for falling out of the party line. The US has changed too and things that would have been unthinkable during the filming of the documentary are now reality.
3. I want to be an entrepreneur of a tech/robotics company. I think Ai won’t take this job because starting a company isn’t one clean task that a model can do end to end. However, AI can supercharge an entrepreneur’s productivity. Jobs from 1950 that don’t exist now are telephone switchboard operators, elevator operators, and human “computers” (from Apollo missions and such). Jobs from 1990 that are gone are video rental store clerks, one-hour photo developers, and travel agents for regular trips (though a small luxury and corporate niche survived mainly for older companies/people). Jobs from 2020 are harder because it’s only been six years with a lot of these jobs not being fully extinct but just hollowed out. Manual transcriptionists, translators, and SEO optimizers are some examples I can think of.
4. Cyberattacks on infrastructure are probably the biggest threat. Almost everything runs on computer systems now, power grids, water treatment, hospitals, banks. AI makes it much easier to find weaknesses in those systems and attack them. A smaller country or even a terrorist group with no country at all could do serious damage to a much bigger nation, and it’s hard to trace back to them. Autonomous weapons are another concern, along with where the chips actually get made. Most advanced chips come from Taiwan, which is also the place the US and China are most likely to fight over. That means the entire industry depends on a very unstable location. The key players are obviously the US and China in general AI. Additionally, India will be a key player because it has what China is losing, a young population instead of an aging one. It also has 1.4 billion+ people generating data, and a huge number of software developers. Its economy is rapidly growing and its people have some of the highest comfort levels with AI of any population in the world. Japan is also important for a different reason, physical AI. They’re already scoring ahead of the US in recent robotics indices, and they lead in industrial robotics with companies like FANUC and Yaskawa. Once progress on LLMs plateaus, I think physical AI becomes the next big shift, and Japan is positioned to play a key role in it. They also have a real need for it, since they’re a super-aged society without enough workers. Culturally, Japan sees robots as helpful rather than threatening, which is the opposite of how a lot of Western pop culture treats them (Terminator versus Doraemon) which can reduce a lot of cultural friction with adoptions. Another place Japan could play a big role is chips. Before the rise of Taiwan in the 80’s, they dominated semiconductors, with them still having the largest number of semiconductor fabs. Some companies there are already making factories that can make 2nm chips and EUV fabrication at the same plant. The US will also benefit with new chip manufacturing as TSMC seeks to move its center of production out of Taiwan. For the diplomatic side, I think the EU ends up aligned with the US, even though they want independence. The EU’s problem is that a lot of their AI research is top notch but non commercializable and they have too many regulations. Russia will probably end up aligned with China, mostly because of sanctions, but they’ll likely stay strong specifically in military AI given how good their defense tech already is. Overall I think they’ll be two blocs: US, EU, India, and Japan on one side, and China and Russia on the other.
5. Overall I think these tools have been a force for good. They basically put a tutor in everyone’s hand, something that wasn’t accessible a few years ago except to the richest people. Before this, if you wanted one-on-one help whenever you got stuck, you needed a private tutor, and most people couldn’t afford that. For computer science specifically, I think AI has been good for me because I’ve never really liked syntax and formatting. What I actually care about is the ideation part, coming up with something and figuring out what CS can do to solve a real problem. AI helps me get to that part faster instead of getting stuck on syntax. That’s what made me want to learn CS in the first place, the ability to build something that actually does something, not memorizing exact syntax rules. That said, I don’t think it should replace basic thinking. I see it kind of like a calculator. Everyone uses one now, but kids still learn arithmetic first, and I can do math without a calculator even though I use one constantly. I think coding needs to work the same way. AI can write the code, but you should still understand what the code is doing and why, otherwise you can’t tell when it’s wrong or fix it when it breaks.
I really liked your analogy with Pearl Harbor and oil compared to data collection today. I’m certain that data will get to a point where countries will have conflicts over who has access to which datasets; espionage could go from stealing secrets to stealing large amounts of data that can help train a model to be more ‘intelligent’. Future conflicts over data will likely not be conventional; instead, the battle will likely involve technology, bugs, etc.
I also think the film is pretty out of date—lots of things that were deemed unacceptable or unimaginable in the video are reality today. Being upfront about data collection is a key step toward the American public accepting it, but progress will inevitably be slower than in China, India, or Russia.
Finally, I think that using AI to boost productivity and ability will be something valuable in the near future. AI can remove the necessity of grunt work, so hard work no longer opens as many doors as it used to. If you can master AI and use it to improve your productivity, doors are still open for you, and the potential grows exponentially.
I found your answer to Question 4 really interesting — I didn’t know about a lot of the details you brought up about Japan in particular, and I never considered cultural differences in attitudes towards robots as being a factor in advancing AI (although it makes a lot of sense).
On your answer to questions 4, it was really interesting to learn about the Japan robotics stuff because I get no news about foreign robotics innovations. Also in response to your answer to question 4, I don’t think that India will be on the US-side, because of strong India-Russia military trade and collaboration, and weak India-US relations with Trump’s second term.
I was interested reading your prediction of what countries would play key roles in the future. You made a good point that regulations can play a big roles in the growth and monetization of AI. I would agree that China has the biggest advantage in terms of data amount especially since the citizens can’t opt out. Data collection built into the culture could give them a great advantage. Ideally there are many more countries than Taiwan and Japan making chips to reduce the chance of war over access to those capabilities. However, all the data collection reminds me of 1984, which we just started reading, where peoples lives are constantly monitored.
Your point about Flock cameras capturing everyday movements without public debate or user consent hits on a key issue in data governance. It shows how surveillance technology often gets used incrementally without clear boundaries regarding who owns, accesses, or searches that location history. Your breakdown of Japans advantage in physical AI and robotics was also insightful.
1. Data is the “new oil” because it’s a resource that the entire world wants but that doesn’t exist in equal quantities globally. China is the “Saudi Arabia of data” because it controls vast amounts of data, giving it political/economic leverage since they control who it is sold to/used by. China’s data ensures its companies have the leg-up in AI, machine learning, and robotics, industries that require vast amounts of data to refine models. By monetizing this data, China profits from selling it, making it a major source of capital. Because China has a data monopoly, they can use it to force foreign companies to comply with Chinese standards, giving them broad control over corporate affairs within the country. Since the documentary was filmed, data collection has vastly improved, allowing companies to provide personalized shopping recommendations and optimize supply chains.
https://www.forbes.com/councils/forbestechcouncil/2025/01/28/how-ai-has-changed-the-world-of-analytics-and-data-science/
https://finance.yahoo.com/news/us-advisory-body-says-chinas-130512629.html?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAJewhXWDDEOk7miFgu5on6hV4_6Li2BHM8tIRYvWXY2wctCCVGGfeKgWE4ZZ_71GlTrDe_kj4pjARpnJYd2D-BRaBBMyc_fL69h_4Nybh4rIlE9F8glwCUcKamw0tYmVAmth5PMq-nrgondGHBELu5LZV7oL46BkAVcBj5UWL_Yt
2. Tracking of U.S. citizens is probably beneficial in the race for AI leadership, just because models improve with more data. I personally think that a lot of the tracking and data collection is an invasion of privacy, or at the very least a sketchy process that should be regulated and clarified a lot better to everyday consumers. As a whole, I think Americans are woefully unaware of their rights, and companies take advantage of that by making disabling cookies difficult. Data collection in other countries is probably not that different, though in China there is no choice to disable tracking and most of it is centralized in the government. It’s still an invasion of privacy, but there isn’t as much of a legal basis to challenge it in authoritarian countries.
3. I looked into operations research analysts, statisticians, and financial quantitative analysts, and all have a lower risk of being automated by AI (around 20-30% chance). 1950s jobs that don’t exist anymore include telephone switch operators, milkmen, and elevator operators. 1990s jobs include projectionists (who projected movies) and encyclopedia salesmen. In the 2020s, translators and entry-level digital operators have also begun to be automated.
https://www.ironmountaindailynews.com/lifestyles/life/2025/04/jobs-from-the-1950s-that-barely-exist-today/
https://www.aol.com/11-jobs-form-90-disappeared-154950738.html
https://fortune.com/article/what-are-the-jobs-most-exposed-to-ai-microsoft-researchers-list/
4. I think my generation’s biggest challenge will be the crisis of meaning as a result of people losing jobs. The vast majority of the population is incapable of high-skilled, PhD level labor, so leaving potentially billions without a job will most likely trigger a global crisis where people—now unable to work and earn money—turn to depression and violence. At the very least, governments will have to find ways to fund the survival of millions who don’t produce value by themselves. In IR, the biggest competitors will be China and the US, each with allies like Russia and the EU respectively. These countries will likely use AI to spread disinformation to manipulate elections and to cyberattack other governments, keeping the military on constant high-alert. I’m not sure that this will dramatically change diplomatic relationships insofar as these countries are already on high-alert and constantly hacking each other; only the magnitude and sophistication of these attacks will increase.
5. AI chatbots have made students feel like skills such as writing are going to become obsolete, making them care less about the process of writing. Simultaneously, teachers have had to change how they teach, doing more in-class assignments to minimize the risk of AI usage. I wanted to learn CS because I like all things logic, and the most interesting part of coding is the underlying math and logic, not the syntax. Besides writing, things like arithmetic are also under threat because students just google AI overview the answer. I don’t think losing those skills is good in any way; even if an AI could get the answer faster, the point of those problems is not the answer, but the process of doing it and practicing through repetition. Yes, you may not use calculus in everyday life, but the ability to think logically through a proof or to visualize the scale of a function’s slope is essential.
On question 4 you mention that governments will have to fund people who won’t earn money. But if AI is able to allow people to live without needing the work because they produce the resources we need in enough quantity, is the freedom not to work not a good goal? In a society where everything is automated wouldn’t it make sense to give everyone free resources? You mention people turning to violence and depression, but in earlier stages of humanity there wasn’t as big of an emphasis on working, just surviving, and surviving didn’t always take a 9-5.
I wonder if a crisis of meaning due to job loss could also have the effect of fueling anti-AI sentiment and creating a kind of movement to encourage more analog activities (I mainly see this in the art community currently). That crisis isn’t something I really considered before, but I do think it would be pretty significant, especially for our generation.
1. He means data is the resource AI runs on, the same way oil powered the industrial economy, and China has more of it because of its population size and looser rules on collecting it. Companies with data monopolies get better models, more accurate ad targeting, and in China’s case a working surveillance system. Since the documentary was filmed the scale has grown a lot, since data now trains entire AI models rather than just recommendation algorithms, and it shows up physically in things like license plate readers and store cameras.
2. I don’t think mass tracking is what determines the AI race, since compute and researchers matter more than surveillance data. Companies and the government should be more transparent about it either way, because cookie banners are mostly theater and nobody reads a privacy policy. The difference with China isn’t the amount collected but who controls it, since here it’s mostly private companies selling ads and there are courts and lawsuits, while there the state has direct access and uses it politically.
3. I’d consider software engineering or data science, and the site rates both as low risk, though its ratings are based on older research that predates modern AI, so I’d guess they’re optimistic. Jobs that existed in 1950 and don’t now include elevator operators, switchboard operators, and human computers, and from 1990 you’ve lost video store clerks, travel agents, and one hour photo developers. Nothing has fully disappeared since 2020, but entry level copywriting, translation, and basic customer support have shrunk noticeably.
4. The biggest challenge is that AI is dual use, so the same systems that speed up research also enable cyberattacks and disinformation, and unlike nuclear weapons you can’t verify what another country has. The main players are the US and China, along with the chip manufacturers everything depends on. I think it ends up as two separate tech stacks with export controls as the main pressure point, with the EU regulating more than building.
5. It’s much easier to get unstuck now, but the tradeoff is that reading an explanation feels like understanding when it often isn’t. I got interested in CS because you get immediate feedback on whether your reasoning was right, which most subjects don’t offer. Losing syntax memorization doesn’t bother me since that was never the real skill, but losing writing does, because writing is how you find out whether your thinking actually holds together.
I agree with your answer to question 5. Using AI to speed up reading processes is definitely a huge time saver and you feel like you are understanding the material based on the brief summarizations from the AI, when in reality if you are asked a question about the AI summary slop that you read, it becomes incredibly difficult to produce real thoughts on the material. Also, your reasoning for being into CS completely resonates with me because I love seeing immediate results of my actions or at least knowing if I am on the right direction instead of going the wrong route.
1. Data is the new oil because the data can be used as a resource for AI models and other algorithms. Companies benefiting from data others don’t have: Google, Google has collected mass amounts of web data over many years. Therefore the company has an advantage in volume data to train its AI. Data collection changed the world because it created the systems to collect data. Things like Cookies, which are active on almost every website could easily be adjusted to take more data, violate the agreements, or be controlled by governments. Data collection has changed the world because with it, specific audiences can be targeted and manipulated for specific purposes.
finance.yahoo.com/news/googles-data-access-gives-huge-201731481
2. I think companies should be very clear about what data they collect. Just like you wouldn’t want anyone to have access to your phone number, any other data shouldn’t be collected without your knowledge. Cookies can be very helpful however because they give you ads for products you actually want. I think the government collecting data is even worse than private entities, the government has too much power to enforce the collection. Cookies add to a surveillance state. Data collection for AI is fine on its own, but the data spreading elsewhere is a problem.
3. One career I’m considering is industrial engineering and it has a 28% chance of being replaced by AI. Pin setters don’t exist anymore in bowling alleys, switchboard operators, railroad signallers, camera film developers.
https://studyworkgrow.com/20-jobs-that-dont-exist-anymore/
4. As private companies gain more and more money I think they will have a greater power over the government. Countries such as the US and China will most likely gain lots of money from AI pushing their economic growth further than other countries. I see China as the main player. China has plenty of unused land and billions of people to build data centers and program new types of AI. As data becomes currency essentially, countries that have more data about their people will be able to make a large amount of profit by selling AI training data to other countries. They can easily sustain themselves. I can see my generation and politics becoming even more focused on money and advancement. I anticipate something similar to the space race.
5. Public AI models have greatly changed the landscape of learning. First of all AI is able to retrieve and organize information in an instant. This makes manual research and writing tasks far less appealing for many students, but as a result they also lose those skills. AI can also be a great tutor and explain subjects like math. I want to learn computer science for problem solving, it allows me to create my own custom solutions to problems. It’s similar to engineering, but with quicker access to testing your project. I believe many writing tasks will be replaced, meeting notes this has already happened for. It can also write papers, so obtaining information and explaining it is important, but formatting and writing choices may have less value when AI can do that.
I think your point about data regulation is interesting. I wonder how much personal data Americans would be willing to sacrifice if it led to the United States remaining the global leader of AI.
1. Kai-Fu Lee meant data is the main resource for AI like how oil was for industry. He says China is the Saudi Arabia of data because in China, data collection is extremely inexpensive and easily accessible. Three benefits are better models, making money off of selling data, and surveillance. Since 2018, there has been a significant increase of synthetic data being used to train AI models. Researchers are realizing that high quality human generated data is running out, and instead are turning to things like simulations. In the case of autonomous driving, Waymo has used over 20 billion simulated miles to learn from, while only having about 200 million real miles. With an increased reliance on synthetic data, I think there are big risks associated, especially because AI systems are already in the physical world and can make real decisions based on conditions only seen on a computer screen.
2. I think tracking has helped US companies lead in AI technologies, but most of the data collection that happens is because of consumer negligence. For most people, when cookie popups appear they click the bright green button that says “I Accept” without understanding what they are accepting to. I think a big difference between data collection in the US versus China is that in China, citizens are protected from private companies spying on you but not the government, and in the US it is partially swapped. I think the issues that come with Flock cameras are very big and not thought about enough as the installation of them increases. In my neighborhood, there have already been Flock cameras installed at each entrance without any discussion within the HOA (according to my dad). I think there should be a lot more clarity about data collection from both private companies and the government.
3. I’m thinking actuarial sciences(21%), software engineering(27%), or Computer and Information Research Scientist(11%). I think the entry level positions are at the most risk, because AI can replace them. However, once at the higher levels, experience is necessary and can’t be replaced by AI. Jobs from the 1950s that are gone are ice and water deliverymen and telegraphers. 1990: pager salesmen, travel agents for stores. 2020: toll booth collectors, entry level translators.
4. I think the main challenge is AI becoming a national security asset like nuclear technology, except that it will be owned by private companies with little government oversight. Key players are the US, China, and EU regulators, as well as Taiwan because of the massive amount of vital chips they make.
5. LLMs make getting the first answer easy, but it makes it hard to actually understand concepts. I take computer science because I like building things with my brain and I have been doing it for a few years. A skill being replaced is syntax memorization and the focus is shifting to thinking speed and creativity. I think this is a good thing because it encourages curiousness, but I do still think that some foundations are needed to point out when an AI can be wrong or you know a better way of doing things given the context.
1. When Kai Fu Lee compared data to oil and China to Saudi Arabia, he meant that data is the fuel powering the AI economy today, and China’s massive population generates more of this than anyone else. In 2026, data monopolies allow companies to build superior predictive algorithms, take out global economic markets by shutting out smaller competitors, and have social control, as seen in China’s social credit system shown in the documentary. This data mining has transformed the physical world through the deployment of smart and (questionably) helpful city surveillance cameras , such as flock cameras, and autonomous logistics, while changing our virtual world by personalizing the media, ads, and information we take in every second.
2. As a user of social media, I know that mass data tracking provides the massive datasets necessary for the US to stay in the AI race, but collecting it secretly is a invasion of individual privacy. The documentary clearly shows that data collection in the US is different from China, as in authoritarian states, data directly feeds government surveillance, whereas in the US, it is primarily taken by private tech companies to increase profits. Because of this, both the U.S. government and private entities should be upfront about how tools like cookies and Flock cameras track us, ensuring we have our freedoms while still advancing technologically.
3. When looking at careers on will robot take my job, roles requiring a lot of human emotion, like nursing or physical therapy, have a very low risk of automation, while highly repetitive roles like accounting and data entry are vulnerable. Historically, automation has always forced large adaptation, for example, switchboard operators vanished after 1950, video store clerks disappeared after the internet became a thing, and toll booth operators and basic cashiers have mostly been replaced by self-checkouts and scanners. However, the documentary emphasizes that the current AI wave is historically unique and threatening because it is replacing cognitive, white collar jobs just as fast as blue collar manual labor.
4. My generation will face challenges as the geopolitical AI race increases, particularly the threat of autonomous weapons, AI driven cyber warfare, and deepfake campaigns used to threaten rival nations. The documentary shows the United States and China as the absolute superpowers in this technogical War, while the European Union focuses heavily on ethical ai regulation and Russian federation states invest heavily in military applications. Ultimately, this race will reshape global diplomacy by forcing smaller nations to choose sides and form AI blocs based on which superpower’s infrastructure they rely on for their economy and security.
5. ChatGPT and Gemini have changed the educational industry by acting as 24/7 tutors, shifting the focus of learning away from simply memorizing facts to applying information. I wanted to learn computer science not just to write basic code, but to understand the logic of algorithms so I can build useful projects with them in the future. Basic memorization and writing will likely be completely replaced by AI which frees people for complex creative problem solving, but we must use human reasoning so we do not lose our ability for independent thought.