Special edition: 4 predictions about AI progress in 2025 — OpenAI prospects, robotics advancements, development automation, and the AI market positions of China and Europe

Viktar Schalenchanka and I tried to predict how AI would develop in 2025. Here are four of our forecasts on the topic.

#1 — OpenAI risks losing its “open soul”

A discussion about OpenAI is inevitable. Currently, the company is striking a balance between being a “technological beacon” and becoming too “corporate” or militarily focused. Google has Gemini, Anthropic has Claude, and ambitious startups in China are gaining momentum. The era of ChatGPT’s unchallenged dominance might indeed be nearing an end. OpenAI risks losing the “open soul” that initially attracted developers worldwide if it continues to actively collaborate with the corporate and military sectors.

#2 — Robots will perform more routine tasks

A breakthrough in robotics awaits us — the creation of more mass-produced androids for routine tasks like cleaning or picking up scattered items. After surging in the manufacturing and service sectors of the economy in recent years, I believe the future of robots now belongs to “robo-housekeepers.” And once developers can integrate powerful language models, users will be able to have dialogues with robots instead of just pushing buttons. There is a flip side, however: the likelihood of accidental or even intentional harm from robots increases with their autonomy. By 2025, the first incidents requiring serious regulatory measures may occur.

#3 — Development automation will increase

By the end of 2025, companies may start hiring people with no training in classical programming. Modern IDEs and “agents” based on GPT-like models may generate code prototypes within minutes. This could edge out specialists who are unwilling or unable to engage with the new technology. But the need for human input won’t disappear — someone has to make strategic decisions, optimize architecture, and ensure final quality.

#4 — China will approach AI leadership, Europe will not

Chinese companies are actively working on large language models and specialized chips. We think that by 2025 they will come close to market leadership. Europe, on the other hand, will not boast a plethora of loud startups due to its strict regulations. Unless it softens its “AI Act” and gives more freedom to innovation, it cannot hope to achieve AI leadership.

AI is not only magic, but it is also constant competition, new risks, and tough choices. If everything happens as we predict, by the end of 2025 we may face a new reality in which the next Netflix hit is generated by a neural network and a talking android takes over cleaning at home. I can’t wait to see this future live, although it’s scary!