Over the past couple of years, generative artificial intelligence (genAI) has rapidly advanced from basic engineering and training adjustments to incorporating external knowledge sources to enhance chatbot responses.
The latest breakthrough in genAI is the introduction of autonomous agents, or AI-powered applications capable of perceiving their surroundings, making decisions, and taking actions to achieve specific objectives. The key element here is “agency,” enabling software to act independently. Unlike traditional genAI tools that focus on generating content like text, images, and music, agentic AI emphasizes proactive problem-solving and complex task execution.
An AI agent can be defined as a combination of a large language model (LLM) and a conventional software application that can operate autonomously to complete tasks.
In a report by Deloitte, it is predicted that by 2025, 25% of companies utilizing genAI will initiate agentic AI pilots or proofs of concept. By 2027, this number is expected to increase to half of all companies. Deloitte suggests that some agentic AI applications may be integrated into existing workflows as early as 2025. These advancements could enhance productivity for knowledge workers and streamline various workflows; however, widespread adoption may take time due to the autonomous nature of these systems.
Major tech companies are swiftly introducing genAI-based agents into their offerings. Microsoft recently announced automated agents for M365 Copilot while Cisco unveiled customer service agents in October. Atlassian introduced its Rovo genAI assistant alongside Asana’s announcement of AI Studio for building agents in the same month.
This indicates that AI agents could soon become as prevalent in workplaces as other genAI tools.
Agentic AI operates in two primary ways: specialized agents capable of independently completing tasks across different platforms such as the open web or mobile apps; conversational web agents function similarly to chatbots engaging users through multimodal conversations beyond simple text chats during web browsing or app usage according to Larry Heck from Georgia Institute of Technology.
Heck predicts that these conversational web agents will empower users with more freedom compared to traditional virtual assistants like Siri or Alexa which are limited within specific ecosystems. He anticipates widespread use cases for AI agents particularly through extensions on search engines and existing virtual assistants like Siri or Alexa.
Other uses…
2024-12-07 17:15:02
Link from www.computerworld.com