From answering questions to carrying out complex tasks: The next phase of AI

Artificial intelligence is evolving beyond simple text, image, and answer generation. At the Google Cloud Summit in Doha, the company highlighted the growing importance of autonomous or "agentic" AI. While generative AI creates new content based on prompts, agentic AI executes tasks by handling the necessary steps to complete a job.

Google AI expert Alex Rutter explained that agentic AI is more focused on streamlining business processes and workflows. One example given was invoice reconciliation, where an AI agent could efficiently check an invoice against multiple sources before approval. However, as AI assumes more responsibilities, human oversight becomes increasingly crucial.

Google advises maintaining a "human in the loop" to verify AI-generated results, though the extent of oversight depends on the task and risk tolerance. In sensitive fields like medicine, AI should assist rather than replace human decision-making. Google Cloud marked three years in Qatar with an expanded national skilling program aiming to reach over 50,000 learners by 2030.

The company is also gearing up for agentic AI by enhancing its data center infrastructure. A new innovation lab, a collaboration between Qatar's Ministry of Communications and Information Technology and Google Cloud, will help test AI projects before real-world implementation. While the impact on the workforce remains uncertain, Rutter anticipates the emergence of "agent managers" who will oversee AI agents alongside traditional human supervisors.