The article outlines the rapid advancements in artificial intelligence, particularly in coding and reasoning capabilities. Since 2022, developments like ChatGPT and Claude have enabled machines to program effectively, automate complex tasks, and even conduct scientific research. The author shares personal experiences implementing AlphaGo using Claude, highlighting its ability to generate code, suggest experiments, and analyze results. This shift means that coding agents can now autonomously conduct experiments, reflecting on outcomes and adjusting methods in real-time, a significant leap from earlier automated tuning systems. An example of Claude's utility is its command structure for running experiments, where it can create folders, execute Python scripts, and save data in organized formats. This automation allows researchers to optimize hyperparameters efficiently, indicate clear goals, and produce comprehensive reports. The article emphasizes that coding agents have evolved into tools capable of tackling diverse digital problems, from developing web browsers to proving mathematical theorems. The author calls attention to the implications of these advancements. With coding assistants becoming increasingly powerful, the potential for automating complex tasks in computer science is vast. This progress raises questions about reasoning and inference, explaining the difference between deductive and inductive reasoning. Ultimately, the focus is on understanding how these tools can change not just coding but the broader landscape of problem-solving in various fields, suggesting a significant shift in how we approach complex challenges.