Sally Guyer chats to Artificial Intelligence Tool ChatGPT, about the use of AI in contracting, the recent ban on the use of the tool in some organizations, and the potential risks for companies that fail to embrace artificial intelligence tools.
Sally: Hey ChatGPT, I would like to conduct an interview with you about the use of AI in Contracting – are you happy to do that?
ChatGPT: Absolutely! I'd be happy to help answer your questions on the use of AI in contracting. Please feel free to start with your first question.
Sally: Great! So, my first question is how can contracting professionals (they might be lawyers, contract managers, commercial managers, procurement professionals – anyone who works with contracts, buy, or sell side, pre or post award) use ChatGPT to help them in the roles that they perform?
ChatGPT: Artificial intelligence, including models like ChatGPT, can be incredibly useful in assisting contracting professionals in numerous ways:
Remember, AI tools should be used as assistants to human expertise, not replacements. While AI can help automate and streamline certain processes, human judgment and expertise are critical, especially in nuanced and complex contracting scenarios.
Sally: In the context of Research and Insights, there was a recent case in the news where ChatGPT made up a legal case that didn't actually exist. Given your propensity to hallucinate and the fact that your training data from the internet only goes up to 2021, would you qualify your ability to provide Research and Insights?
ChatGPT: You're absolutely correct in highlighting this point. As an AI language model, I can generate information based on patterns and data up to my training cut-off in September 2021. However, I can't access the internet or databases in real-time, nor can I verify or update information beyond my training cut-off.
As for the problem of hallucination (generating information that is not accurate or making up facts), this is a known limitation with AI models like GPT-3 and GPT-4. It is crucial to verify any legal case, regulation, or specific fact provided by an AI with a reliable source. Regarding the provision of research and insights, it's important to note that AI can be useful in generating hypotheses, suggesting ideas, and providing general information based on historical patterns and pre-existing knowledge. However, these outputs should be validated with up-to-date and reliable sources, especially in fields like law and contracting, where accuracy and currency are paramount.
The value in AI lies in assisting with tasks that involve pattern recognition, automation, and generating insights based on large amounts of data. But AI should be used as a tool that supports, rather than replaces, professional judgment and rigorous research. This is especially true for contracting professionals, given the potential legal implications of their work.
Sally: Some companies have banned the use of OpenAI’s tools in the workplace. What does OpenAI believe is behind this decision; and is this a concern for the business, as it seeks to expand its market presence?
ChatGPT: While I can't speak directly to the decisions of any particular company as I don't have realtime access to information or updates post-September 2021, I can speculate on some possible reasons why some companies may choose not to use OpenAI’s tools in the workplace:
As for the second part of your question, any company in the AI space should see these concerns as an important part of the business landscape. Building trust with users is critical. Therefore, addressing concerns around privacy, accuracy, bias, and compliance is crucial for expanding market presence.
OpenAI's mission is to ensure that artificial general intelligence (AGI) benefits all of humanity. This means the organization is deeply committed to understanding these challenges and working towards mitigating them. OpenAI has already made significant strides in improving the safety and usefulness of its models, as well as addressing issues of bias and fairness, and continues to actively research and implement ways to improve in these areas.
Sally: What are the greatest risks for business that do not embrace AI in the next 5, 10 or 20 years?
ChatGPT: Businesses that don't embrace AI could potentially face a variety of risks over the next few decades:
The exact risks will depend on the industry, size of the business, and other factors. However, it's clear that AI has the potential to play a transformative role in various aspects of business. Businesses that are slow to adopt AI may find themselves at a growing disadvantage over time. It's important for businesses to actively consider how AI could impact their industry and to develop a strategic plan for AI adoption.
Sally: Thank you ChatGPT, my final question is what tips would you give to contracting professionals about how to work most effectively with you?
ChatGPT: Certainly! Here are a few tips for contracting professionals (or anyone) on how to work effectively with AI models like me, ChatGPT:
Remember, AI is a tool to assist you in your work. It can greatly enhance your productivity and effectiveness if used appropriately and wisely.
Join AI Contracting Week from Monday, 25 September to Thursday, 28 September 2023, where we will explore the transformative potential of Artificial Intelligence (AI) in shaping the contracting landscape.