At a conference in Chicago on Tuesday, Microsoft CEO Satya Nadella informed attendees that the business is training a new set of AI tools to “act on our behalf across our work and life.”
AI developers are marketing the next generation of generative AI chatbots as AI “agents” capable of more beneficial tasks on behalf of humans.
However, the hefty expense of developing and maintaining AI technologies is making many investors wonder if the hype surrounding the technology is unfounded.
A world in which “every organization will have a constellation of agents — ranging from simple prompt-and-response to fully autonomous” is what Microsoft stated last month as part of its preparations.
Such autonomous agents “can operate around the clock to review and approve customer returns or go over shipping invoices to help businesses avoid costly supply-chain errors,” Microsoft explained in a blog post on Tuesday.
Microsoft serves its large business clients with its yearly Ignite conference.
The shift to so-called “agentic AI” occurs as some users perceive limitations in the extensive language models used by chatbots such as Microsoft’s Copilot, Google’s Gemini, and OpenAI’s ChatGPT.
These algorithms do well on some writing-based tasks by guessing the most likely word to appear next in a phrase.
However, for AI tools to access the web, manage computers, and carry out tasks independently on behalf of users, tech companies have been attempting to develop AI tools that are better at longer-range planning and reasoning.
Laptops 1000Marc Benioff, the CEO of Salesforce, has blasted Microsoft’s change in direction.
Salesforce also offers a service called “Agentforce” that employs AI for marketing, sales, and other tasks.
Is Microsoft changing the name of Copilot to “agents”?
In a social media post last month, Benioff stated, “That’s panic mode.”
He continued by saying that Copilot, Microsoft’s premier AI assistant, is “a flop” that leaks company data and is erroneous.
