Part 4/9:
The rapid development largely derives from innovative techniques like reinforcement learning. OpenAI's recent updates suggest that competition among various AI labs pushes the boundaries of what these models can achieve. Both reasoning capabilities and autonomous tool usage are converging, yielding systems that can conduct sophisticated research at speeds outpacing human experts.
Ethan Mollick, an AI professor, articulated that we are now moving into what he calls the era of research, wherein AI systems exhibit reasoning comparable to human expertise. He demonstrated how AI can produce work that typically requires a PhD's input in mere moments, an achievement that most AI models couldn't boast only weeks prior.