10 IN 10: I Researched How AI Is Impacting the Economy So That You Don’t Have To

Let’s face it: we’ve all used AI at some point: whether it was to copy the Math homework from ChatGPT 5 minutes before class, make a fun video with our friends or just by searching something in Google. But how did it all happen so fast? AI was once something they wrote about in sci-fi graphic novels and now I can’t reply to a two-sentence email without asking a chatbot to draft a kind and friendly response. What changed? 

If you find it hard to keep up and find this new world confusing, you’ve come to the right place. The world of AI, which is expected to increase the global GDP by 7 percent by 2033, is ever-evolving, so if we’re honest, it’s practically impossible to keep up with every new development. There’s probably a new advancement as I’m writing this right now. In fact, if you have the bad habit of having FOMO (Fear of Missing Out), learning more about AI is good practice, as you’ll find it frustrating to attempt to catch up. You’re not alone: a study that asked people to answer (yes or no) if they have a good understanding of what AI is, the number of people who say ‘yes’ increases from 2022 to 2023 but plateaus from 2023 to 2024. 

Artificial Intelligence is the fastest evolving technology in human history. It simulates (or at least tries to) human intelligence by recognizing patterns. The best example of this are LLMs (Large Language Models), which can predict the next word because they have been ‘fed’ thousands of pieces of data on human behaviour. For instance, if I input ‘twinkle twinkle little—’, it will output ‘star’, as it has analyzed the probabilities of that being the next word and it was the highest. However, over time, it has reached many levels of achievement, and some even train themselves now. It could potentially impact millions of jobs and, depending on who you ask, jeopardize—or encourage—economic growth. 

The first question I’d like to address is if it’s profitable. To do this, I accessed a website called ‘Is AI profitable yet?’, which analyses all of the big-tech companies' investments and their ROI (Return on Investment). Most people believe that because it’s shiny, new and making all the headlines, the industry is generating millions of dollars. That couldn’t be less true. At the time of publishing, total industry spend is 1.5 trillion and total industry revenue only 769 billion dollars. Most haven’t even broken even! In fact, I decided to put this to the test. Since the page loaded, it took 34 seconds for tech companies to spend 1 million dollars, which is what buying a 4-bedroom villa in Bali costs. In 34 seconds!!! That’s how crazy we’ve become.

According to Yahoo Finance, OpenAI (the creators of ChatGPT), will be profitable in 2030 at earliest. Just the hardware part of the industry is generating revenue right now. The only profitable in fact are Micron, Nvidia and AMD. Unsurprisingly, they take care of supplying chips to big-tech companies like Anthropic (creators of Claude), OpenAI and Microsoft (creators of Copilot), among others. To be honest, they’re the smartest, as they’re quietly making billions while the other tech giants struggle in negative revenue. 

Why is this happening? A very common misconception of AI is that it’s something that only exists digitally, when all the information is actually stored in data centers that take up huge amounts of power. Those data centers weren’t here a few years ago, so tech companies need loans in order to build them, which is perfectly understandable. The problems start when AI use keeps increasing but willingness to pay for the services decreases every day. This is known as price-sensitive. Let’s look at ChatGPT. Out of its estimated 900 million users, only 50 million have subscriptions, roughly 5 percent. 

Many investors but also ordinary people are putting their money in, but it’s not coming back. This could lead to so many issues we have seen in the past, like tulips in the Netherlands in the 1630s, the Internet in the 90s and Bitcoin in the early 2010s. Sound familiar? It’s known as the AI ‘bubble’, which is when the amount of money going into an industry is a lot more money than what is coming out of it, and any day now it can burst, causing job displacement and bringing down the whole market with it. Think of it like a needy friend who constantly borrows money, promising that they’ll give back more, when they never even give the original quantity back. Proof of this is that large tech companies are already asking the American government—where there is a general lack of expertise of the industry—for funding, which is always a red flag that signals desperation.

All of this is heightened by the fact that productivity gains from AI are below expectation. AI must profitably replace 10 million white collar jobs (out of 70 million) in the US by 2030. Nevertheless, in an ‘Asian Dad Energy’ video, a former software engineer with 25 years of experience in the tech industry says it will only replace 100 thousand in the same period, optimistically. 

An article by Goldman Sachs claims that 7 percent of American jobs will be substituted, 63 percent complemented and 30 percent unaffected. It affirmed that ‘many workers that are displaced by AI automation will eventually become reemployed’. That’s something specific about the article that kind of confused me, so I ran the numbers. 7 percent of substituted jobs equals 11.7 million. The unemployment rate in the country is currently at 4.2 percent, roughly 7.094 million. So if the claims are true, how is it possible that 11.7 million people are going to be reemployed in the following years, when there are already 7.094 million unemployed, while AI automation keeps increasing, in spite of it being below expectation?

Sam Altman, the CEO of OpenAI since 2019, is also something I’d like to discuss. Our entire economy is dependent on a man who said in an interview, and I quote, ‘we might screw up’, when referring to AI. Well, thanks for being so reassuring. It’s not like the future of the generations to come depend on your trust. He wants our money, water, electricity, data…without returning anything. Not even the feeling that we’ve made a good choice by trusting him. He wants 250 gigawatts capacity by 2033 which equals 20 to 25% of the US electrical grid capacity. 

Unfortunately, studies show AI is biased and has racist and sexist tendencies, which make the claims that it will build a brighter, more equal future seem questionable. For example, if you ask a chatbot to generate pictures of scientists, most of them will be white and male. This is not a coincidence. In fact, if it’s like this it is because it’s been trained largely by white men. If you look at a global AI talent representation (2016-24), 69.46 percent are male and a mere 30.54 percent are female. 

One of the places in the world where AI has flourished the most is California, enhanced by the presence of some of the industry’s most well-educated engineers from top universities like UC Berkeley and Stanford. Plus, the historical ecosystem of Silicon Valley is already there. Moreover, there is a lot of venture capital, defined as ‘money provided by investors to young startups and small businesses that have high potential for fast growth’. This attracts young entrepreneurs.

The new Cold War between China and the US is no joke. But they’re not fighting for nuclear power or the space race we saw between the States and the Soviet Union in the 1960s. They want something else: control over AI. However, it feels like they’re playing two completely different games. While the US is trying to create high-quality technology that performs at its best in many different areas, China is making cheaper and simpler AI models and integrating them everywhere in society: military, agriculture, surveillance… This resembles the iPhone, designed in California but assembled in China. Although the latter are not innovative, they are also benefitting from something they have not created themselves. 

The controversial yet groundbreaking DeepSeek V-3 model was released in September 2024. It only cost 5.6 million dollars to train in just 55 days. This proves that AI and politics, like everything, are heavily intertwined and influenced by each other. 

AI has also raised many concerns about legal issues on intellectual property, equal opportunity, civil rights, antitrust and consumer protections. There are no regulations and no accountability. Consequently, most of my research for this article was from second-hand sources, and I often found it hard to find updated information. I’d like to highlight the importance of the government imposing laws for tech companies to give sufficient data-based information to correctly understand the impact of evolving technology, instead of biased, misleading information, especially if the industry constitutes up to 40 percent of the US stock market. Awareness and economics go hand in hand, and a lack of the first could result in a downfall of the latter.

Companies have adapted AI in many different ways, adjusting to the products and services they offer to maximise efficiency in the workplace and increase customer satisfaction. 

DuoLingo uses Artificial Intelligence to tailor language learning to the users, highlighting weaknesses to improve while limiting activities on topics that are easier for them. This stops the consumer from feeling bored or overwhelmed. The GPT-4 model (which is a bit like the engine behind ChatGPT, in this case, the car) is used to explain answers to premium users. AI is also used inside the app to track consumer behaviour and send notifications based on recurring patterns. For some, this may result in motivation. For others, it may become dreadful to even open the app because of the stalking bird sending constant messages. Although notifications increase retention by 3 percent, not all users are exactly happy with this form of emotional manipulation.

JPMorgan Chase intends to use AI as a way to leverage time for high-value activities instead of repetitive ones, with the intention of increasing morale among employees. In addition, because fraud detection could cause financial institutions as much as 58.3 million by 2030, the largest bank in the United States will use AI technology to avoid fraud-related losses. 

Amazon already uses AI to personalize product recommendations, forecast demand and optimize inventory. What’s crazy is that suppliers now negotiate with AI instead of vendor managers at the company, like it was a few years ago. This innovation is based on the assumption that humans and their needs are predictable—and taking advantage of it. 

Have you ever wondered why you’re so addicted to Netflix? The streaming service applies AI algorithms to keep up customer retention by analysing behavioural patterns and personalizing the TV shows suggestions, made specifically for you. 

Last but not least, I’d like to discuss how AI is going to impact marketing positions within companies. In a YouTube video, the Chief Marketing Director of Superside, Jen Rapp, explains that when they have a concept for a campaign, they now discuss six different ideas, create a storyboard and approve it, much quicker thanks to AI. It drastically reduces the time it takes to put on ads. But she also said that because it’s easier, there are higher standards to meet. In a TED Talk called ‘How AI Is Changing Marketing’, Jessica Apotheker warns of the risk of a company losing brand identity. She also raises an important question: if AI can do part of my job, do I have more free time in the office, get paid less or work even harder to publish more ads with more ideas, which could lead to content overload? 

Like most things, we, as ordinary people reading an incredibly long blog article, cannot change things. We cannot stop them. We cannot control them. But if you made it this far, then congrats, this means you have what it takes to be aware. Fear is fueled by lack of information. Fear leads to inaction. Fear is dangerous. Being aware, on the other hand, is one of our greatest achievements as young people. Call me naive, but maybe there’s a spark of hope somewhere. As someone incredibly wise—yet terrifying—once said, hope is the only thing that’s stronger than fear. So what’s it going to be? Will we be part of a world where robots with no hearts dominate human beings with no agency?* Or…

*Quote from Van Jones TED Talk in 2026, ‘The Kind of AI We Actually Need’.