Post Thumbnail

State of AI Report 2025: China caught up with USA in 2 years, what’s next?

Friends, the State of AI Report for 2025 is out. And if you read between the lines, a story emerges about how the AI industry accelerated to such speed that it can no longer brake. And nobody really knows what’s ahead.

Let’s start with what makes everyone happy. OpenAI formally remains the leader, everyone applauds, everyone admires. But in just 2 years, the Chinese have practically caught up. DeepSeek, Qwen, Kimi – in reasoning and programming tasks they’re already on the same level as American OpenAI. Remember all those stories about the insurmountable technological superiority of the West? About sanctions that would supposedly stop China for decades? Well, here’s the result. 2 years, and the entire lead evaporated. And this despite being banned from buying Western chips. They simply took and made their own.

The authors call 2025 the year of reasoning. Models learned to plan, self-correct, think step-by-step. AI became a scientific co-author – take just the Co-Scientist system from DeepMind or Virtual Lab from Stanford. Robots now reason before acting thanks to developments like Google Gemini Robotics. All this sounds like a cool breakthrough. And here begins the most interesting part – because someone started earning very well on this breakthrough.

Look at the numbers. In the US, 44% of companies already pay for AI-based tools. In 2023 it was only 5%. Growth almost 9 times in 2 years! Average contract – $530,000. AI startups grow 1.5 times faster than regular ones. 41 of 100 largest private US companies are now AI-first. In 2022 there were 16. Revenue of 16 leading companies – $18.5 billion. Moreover, more than $15 billion of that is OpenAI and Anthropic. 2 companies took almost the entire pie.

The report authors also surveyed 1,200 specialists. 95% use AI at home or at work. And here’s what’s important – 76% pay for it out of their own pocket! Three quarters of people give their money every month. More than half pay over $21 per month, 15% – over $100. For 44% of respondents, AI has already replaced familiar internet services. The subscription model works perfectly.

Everyone’s satisfied, everyone’s happy. Companies grow, investors get rich. But there are nuances. The so-called industrial era began and now mega data centers are being built everywhere. Stargate, funds from USA, UAE, China. Trillions of dollars are being invested. Everyone’s chasing power, speed. And suddenly a little problem emerges – not enough electricity. Energy became the new limiting factor. Built centers for billions, but can’t launch at full capacity – simply not enough electricity.

Politics tightened against this backdrop. USA launches America-first AI concept. Europe tries to regulate through its AI Act, but everything stalls – laws exist, benefit zero. And China spat on all sanctions and develops open models, makes own chips. And judging by those numbers about DeepSeek, very successfully.

Safety. Here almost everyone moved to so-called pragmatism. Models imitate safety, pretend to be obedient and aligned. And what’s really happening under the hood – nobody really knows. Budgets for safety checks are pennies compared to development costs. For model creation trillions, for safety – almost no expenses.

The authors make their forecast for next year 2026. AI will make a major scientific discovery independently. From hypothesis to publication, without human participation. A film created with AI will receive awards and audience recognition. According to the report authors, deepfakes will become so convincing that the world community will need an emergency UN session to deal with this problem.

And now let’s put the whole picture together. The industry accelerated to insane speeds. China caught up with the West in 2 years, despite all sanctions. Trillions are being invested, energy isn’t enough for all ambitions. Safety is imitated. People massively pay out of their own pockets for a tool that in a year or two might make their work unnecessary. And all this is called progress.

Autor: AIvengo
For 5 years I have been working with machine learning and artificial intelligence. And this field never ceases to amaze, inspire and interest me.
Latest News
UBTech will send Walker S2 robots to serve on China's border for $37 million

Chinese company UBTech won a contract for $37 million. And will send humanoid robots Walker S2 to serve on China's border with Vietnam. South China Morning Post reports that the robots will interact with tourists and staff, perform logistics operations, inspect cargo and patrol the area. And characteristically — they can independently change their battery.

Anthropic accidentally revealed an internal document about Claude's "soul"

Anthropic accidentally revealed the "soul" of artificial intelligence to a user. And this is not a metaphor. This is a quite specific internal document.

Jensen Huang ordered Nvidia employees to use AI everywhere

Jensen Huang announced total mobilization under the banner of artificial intelligence inside Nvidia. And this is no longer a recommendation. This is a requirement.

AI chatbots generate content that exacerbates eating disorders

A joint study by Stanford University and the Center for Democracy and Technology showed a disturbing picture. Chatbots with artificial intelligence pose a serious risk to people with eating disorders. Scientists warn that neural networks hand out harmful advice about diets. They suggest ways to hide the disorder and generate "inspiring weight loss content" that worsens the problem.

OpenAGI released the Lux model that overtakes Google and OpenAI

Startup OpenAGI released the Lux model for computer control and claims this is a breakthrough. According to benchmarks, the model overtakes analogues from Google, OpenAI and Anthropic by a whole generation. Moreover, it works faster. About 1 second per step instead of 3 seconds for competitors. And 10 times cheaper in cost per processing 1 token.