The far-reaching impact of ChatGPT-4 for complex tasks

12.2% more tasks completed

25.1% faster task completion

40% higher quality outcomes

These are the findings of a study conducted by Harvard Business School and Boston Consulting Group. Researchers demonstrated that consultants using ChatGPT-4 outperformed their peers who didn’t have access to ChatGPT-4 on 18 different complex tasks.

The study, involving 758 consultants, evaluated performance on 18 tasks requiring creativity, analytical thinking, and persuasive communication.

Tasks included conducting a market analysis, drafting press releases, and convincing colleagues of a product strategy through detailed memos.

This study underscores the transformative power of language models like ChatGPT-4—a finding echoed by the legal professionals we train to integrate AI into their practices.

Additional findings:

• When GPT-4 was unsuitable for a task, consultants performed better without it (84%) compared to using GPT-4 (60%-70%). GPT-4’s convincingly incorrect answers can mislead users unfamiliar with its limitations.
• There’s a significant gap in understanding the capabilities and constraints of GPT-4, making it challenging for users to identify its strengths and weaknesses for specific tasks.

ChatGPT offers incredible potential for handling complex legal tasks. However, as Harvard’s research shows, effective and secure use requires clear guidance. Our legal-focused training provides professionals with the tools to unlock ChatGPT’s full potential safely.

Contact us today to discover how we can help you harness the power of ChatGPT in your legal practice.

Read the full research paper here: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4573321