Photo credit: www.kiplinger.com
Generative AI captured significant attention last year, but emerging trends suggest something even more transformative on the horizon—beyond merely creating text or art. During Nvidia’s GTC conference, CEO Jensen Huang articulated that the rise of AI agents could represent “a multi-trillion-dollar opportunity.”
What he highlighted is a shift away from traditional artificial intelligence, which primarily focuses on content creation, toward AI systems capable of managing complex tasks, automating processes, and functioning similarly to human coworkers.
Huang humorously remarked, “The IT department of every company is going to be the HR department of AI agents in the future,” pointing to the evolving role of AI within organizations.
What is an AI Agent?
AI agents are advanced software systems designed to perform tasks independently. They can perceive their environment, learn dynamically, and utilize various software tools.
According to Aaron Chaisson, Vice President of Product and Solutions at VAST, an AI agent distinguishes itself from a typical chatbot. “While you can input prompts to receive responses from a chatbot, an AI agent operates autonomously, capable of planning and executing multiple steps to reach its objectives.”
For instance, if you were to express a desire to travel to London using a Visa card, one AI agent might handle your flight bookings while another secures your hotel accommodations. A third would suggest restaurants and attractions to visit. However, it’s important to note that full autonomy isn’t implied; human oversight and approvals will still be necessary.
The Leading AI Agents
Identifying the top AI agents remains a challenge, given that the field is still nascent.
“Currently, AI agents are not production-ready for most applications,” noted Andy Triedman, a partner at Theory Ventures. “Many tasks are too complex for today’s models, and while they can handle some tasks, success isn’t guaranteed across the board.”
Nonetheless, companies such as OpenAI, Anthropic, Microsoft, Amazon, and Meta are continually improving their AI systems, making them increasingly agent-like.
OpenAI’s Operator platform, for example, is capable of navigating websites, filling out forms, placing orders, and scheduling appointments. Currently, it remains in a research phase and is accessible through a subscription model, with plans to integrate it into broader offerings.
Furthermore, prominent software firms like Salesforce, ServiceNow, Oracle, and Microsoft are advancing AI agents to automate business functionalities. These companies are well-positioned due to their vast data resources and operational workflows, enabling them to train AI agents more effectively.
Salesforce has introduced Agentforce, facilitating quick development and deployment of AI agents, sometimes within minutes. Their marketing efforts highlight the initiative’s relevance, with notable figures like Matthew McConaughey featured in its commercials.
“Agentforce transcends traditional chatbots and copilots, employing advanced reasoning to make decisions and execute actions, such as resolving customer inquiries,” stated Kishan Chetan, Executive Vice President and General Manager of Service Cloud at Salesforce. He further explained that the Sales Coach Agent aids sales professionals by offering customized practice scenarios, providing feedback, and equipping them with strategies to enhance their sales techniques.
Numerous startups are also pushing the boundaries of AI agent technology:
- MindCorp: Founded by Michelle Crames, a former McKinsey & Company analyst, MindCorp aims to streamline management consulting processes, turning tasks that typically span weeks into mere hours.
- Sierra: Co-founded by Bret Taylor, previously Facebook’s CTO, Sierra focuses on using AI agents for customer service, boasting clients like SiriusXM and SONOS. The company raised $175 million at a valuation of $4.5 billion last year.
- Hippocratic AI: This startup employs its AI models to facilitate healthcare tasks, such as pre-operative evaluations and appointment scheduling, recently securing $141 million in funding at a valuation of $1.64 billion.
The Risks of AI Agents
The capacity for AI agents to make autonomous decisions introduces significant risks. For instance, a system might misread data and authorize an excessive invoice, or a healthcare AI could misdiagnose a condition based on biased training data, prompting incorrect treatments.
“To ensure accuracy and minimize the risk of catastrophic errors, AI agents must be continually updated with the latest information,” emphasized Chris Moss, General Manager of Small Business Solutions at Dun & Bradstreet. “It remains essential to have human oversight in place.”
Privacy concerns also warrant attention. If you’re utilizing AI technology, it’s critical to safeguard your personal data and remain informed about how your information is being handled.
Another significant challenge lies in current IT infrastructures, which are not yet equipped to handle the sophisticated requirements of AI agents. “AI agents must effectively integrate with diverse applications, operate in secure environments, and provide analytic insights to support developers in troubleshooting,” noted Triedman. While progress is being made, complete solutions are still under development.
Learn More About AI
Source
www.kiplinger.com