Tech megacaps ($200 billion or more in market share) plan to spend more than $300 billion combined on AI in 2025, what does this mean in the grand scheme of the AI race?

4 of the largest tech companies in the world-- Meta, Amazon, Alphabet, and Microsoft, intend on spending up to $320 billion combined in order to keep up in the AI arms race. This $320 billion figure would be a substantial increase from the spending figure of these 4 companies in 2024, which was $230 billion. This affirmation of the worthiness of AI comes at an interesting time, as just before this display of confidence came a point of fear: the release of China's Deepseek AI. The timing of this backing shows that American companies are not going to blink in the face of competition; however, I believe that this race to the top raises a few questions. 

The first of which being who is the ultimate winner of the AI race? I think that an interesting place to turn to for a possible answer or way of thinking about this is the Jevons Paradox. The Jevons Paradox states that as costs to use a resource drops (efficiency rises), demand goes up, and so too does total consumption. The reason why I think that this is a good starting place to form thoughts on the issue is that Microsoft's CEO, Satya Nadella, tweeted out using the paradox as evidence that people won't be able to get enough of AI as it gets cheaper and more efficient. However, Jevons paradox, as applied to this scenario, brings about even more questions. Who is the one making AI cheaper and more efficient to use for consumers? The current answer to that looks to be Deepseek and China. Next, what company benefits from total consumption going up. In my opinion, that would be Nvidia, as more and more people using a more efficient version of AI would also lead to Nvidia benefitting from selling their GPU's to China. Continuing on, what companies could get left behind here then? I think there is solid evidence, by using the lens of Jevons Paradox, that could show that ChatGPT and OpenAI, as well as those who rely on those companies are the ones getting left behind here. 

Source for the numbers: https://www.cnbc.com/2025/02/08/tech-megacaps-to-spend-more-than-300-billion-in-2025-to-win-in-ai.html