The quality of researchers, their ability to learn from data, and the efficiency of experiments are crucial inputs to AI R&D progress.

In our model, we expect that by 2027, AI agents will be able to autonomously conduct AI research, learning from their experiments and improving their methodologies through online learning.

As we approach mid-2027, the key factors in AI research will be the taste of AI agents and the amount of compute available for experiments, as these will determine the effectiveness of automated AI research.

While AI isn't conducting new scientific research, it enables scientists to complete their work more quickly and effectively.

The potential for AI to synthesize knowledge and facilitate innovative solutions could lead to a new era of breakthroughs in various fields.

To innovate beyond current knowledge, AI systems must learn to identify patterns across different domains, a capability that they currently lack, which is crucial for future advancements in science and technology.

The emergence of AI agents allows models to autonomously retrieve information, run code, and perform multiple tasks without human intervention, leading to significant economic impacts.

Emerging technologies in AI, such as browser use technology, will allow AI agents to operate across diverse systems, enhancing their ability to gather information and perform tasks that previously required human intervention.

In the coming years, we will see more experiments on how to effectively deploy and manage multiple AI agents for complex tasks.