The closure of startups, job losses, and rising prices for ChatGPT: what experts predict for the world with AI by 2030

In the joint project by Financial Times and MIT Technology Review — “The State of AI” — experts discuss how this technology is transforming the economy, the labor market, and the global balance of power — and why the benefits of AI are distributed unevenly.

Will Douglas Heaven, Senior Editor for AI at MIT Technology Review

Tim Bradshaw, Global Technology Correspondent at FT

A Look at the Future with AI

Will. There are significant discrepancies in current forecasts regarding the development of AI.

One group of experts — the AI Futures Project led by former OpenAI researcher Daniel Kokotailo — believes that within the next 10 years, AI will have an impact greater than that of the Industrial Revolution. In their AI 2027 scenario, they describe possible development paths: from rapid technological growth to deep crisis. The main point is that AI has the potential to radically change the economy and society.

On the other hand, the team from Normal Technology at Princeton holds a different view. Researchers Arvind Narayanan and Sayash Kapoor believe that technologies rarely change the world instantly. Usually, they are implemented gradually, and their real impact becomes noticeable only over time. In their opinion, AI is important, but its effect on everyday life will grow slowly.

Tim. The key difference between companies will not be the speed of creating new models, but how exactly they will apply AI. “Browser wars” are already underway, and composable chatbots are emerging. In the coming years, models will become cheaper and more accessible, and the focus will shift to new ways of using them. At the same time, AI development is not limited to large language models: reinforcement learning and world models — systems that better understand the physical world — are actively developing.

By 2030, the world will change significantly. The AI revolution will develop rapidly, but not everyone will benefit. By the end of the decade, the AI bubble may burst: some startups will close, many specialists will lose their jobs, and users will face rising prices for services like ChatGPT. At the same time, multi-layered use of AI will lead to increased costs — for reliability and security, it will be necessary to run multiple AI agents simultaneously, which will sharply increase computing expenses.

The Accessibility of Future AI

Tim. Physical AI is also actively developing. Robotaxis will appear in major cities, and humanoid robots will eventually begin to be used in homes. However, due to the high cost of computing, maintenance, and infrastructure, these technologies will be available mainly to affluent users in the coming years. For others, AI will remain limited — in the form of simpler and less effective tools.

Will. The gap between those who have access to advanced AI solutions and those who do not will only widen. Already, paid subscriptions to ChatGPT or Gemini for $200+ per month provide a fundamentally different level of capabilities compared to free versions. As costs grow, developers will intensify this divide.

Furthermore, global inequality persists: in countries of the Global North, AI is being implemented rapidly, but where there is no stable access to electricity and the internet, the benefit from AI remains minimal.

About the Future

AI will develop not only in Silicon Valley. Open and cheaper models are spreading quickly, and countries of the Global South are already actively using alternative technologies, including Chinese ones.

Even if the AI boom in Silicon Valley weakens by 2030, the global race for influence over technology — and for who will benefit from it — is just beginning.

Source: TechnologyReview