Urgent.News

What's breaking now, across thousands of outlets.

AI

Microsoft retools Copilot with coding, AI agent capabilities

The software giant aims to capture more of the enterprise AI market as providers such as OpenAI and Anthropic snatch up business, a Gartner analyst told CIO Dive.

Microsoft has unveiled a new vision for its AI assistant Copilot, aiming to integrate it as the entry point to its suite of productivity software, including Word, Excel, and PowerPoint. This integration allows Microsoft 365 subscribers and businesses in its testing program to create, edit, and share files in Copilot without switching applications. Jacob Andreou, an executive vice president overseeing Copilot, confirmed this development to AFP.

This move follows similar initiatives by other AI providers. OpenAI and Anthropic, creators of ChatGPT and Claude respectively, have launched their own document creation tools recently. Both are major suppliers to Microsoft, as Copilot primarily runs on their models. Microsoft introduced its own models in June, but they currently power only a few products like GitHub's coding tool Copilot, according to Andreou.

The company also announced a low-cost model for routine tasks in June, which has yet to launch. Copilot now includes Autopilot, an AI software capable of handling automated tasks independently, previously known as Scout. This tool is limited to a few clients for now. In June, Microsoft described Autopilot as the company's first "always-on" AI assistant that can operate online even when a computer is turned off.

CEO Satya Nadella compared the launch of these AI tools to the introduction of personal computers, stating they could transform the way businesses operate. He emphasized the challenge of enabling companies to transition between AI models without losing data or knowledge. Nadella highlighted the importance of allowing companies to switch AI models without losing their data or expertise.

For a Copilot subscription costing US$30 per user per month, users will access basic AI features such as data analysis and document drafting. Advanced agents, requiring more computing power, will be billed based on usage. Nadou advised companies to identify their top 10 to 20 percent of high-performing employees and grant them access to more powerful AI tools.

Written by urgent.news from Free Malaysia Today's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

This story

This is one outlet's version. Read the fullest account.

Read the original at ciodive.com →

More in AI

5 Questions you would ask your operations team to create better ML models:

1. At what point of time in the designated process, will the team utilize the Model? Understanding the timeline is very crucial for the model development process because it defines the prediction…

  • When in the project timeline will the model be employed?
  • What format is required for the model's output?
  • How frequently will the model's predictions be used?

An Overview of Basic Gradient Descent Optimization Algorithms

Gradient Descent is a method for unconstrained mathematical optimization . It is a first-order iterative algorithm for minimizing a differentiable multivariate function.

  • Batch Gradient Descent computes gradient using entire training dataset for one update.
  • Stochastic Gradient Descent updates parameters with each random training example.
  • Mini-batch Gradient Descent balances accuracy and speed using small training batches.

More from Friday 25 September →