Leading AI developers like OpenAI and Anthropic are racing to create autonomous AI agents for office tasks, a trend highlighted by The Information. As IT transforms into “HR for neural networks,” large language models (LLMs) are evolving to perform complex corporate duties. Traditionally, AI training was two-fold: data consumption and fine-tuning. Now, LLMs learn by operating within virtual office environments, akin to flight simulators. Startups like Turing are already developing training systems mimicking popular applications such as Salesforce and Gmail. This approach enables AI to handle multi-step tasks, marking a shift from generating text to driving real autonomy. While concerns about job displacement arise, experts suggest AI should augment roles, allowing humans to focus on complex issues. The future may see entire economies functioning as RL environments, where AI learns from real-world workflows across various professions. As firms invest heavily in this technology, they aim to enhance productivity and improve workplace dynamics.
Source link
