Disclosure: I am not a lawyer, and do not practice law.
At ipCapital Group we are early adopters on new technology by necessity. As invention, innovation, and intellectual property consultants, it’s crucial to our and our clients’ success that we are intimately familiar with the most promising technology so we can develop business strategy to navigate the changing landscape.
Enter ChatGPT.
As we move towards an increasingly digital world, knowledge work is becoming more important than ever before. From legal research to technical analysis, intellectual property firms are grappling with how to stay ahead of the competition. In this environment, ChatGPT, an AI language model powered by OpenAI’s GPT-3.5 (and GPT-4 beta) architecture, is poised to transform the industry. Firms that know how to effectively deploy ChatGPT in their business will easily outcompete companies who are resistant to adopt. In this newsletter, we will explore best practices for using ChatGPT in the intellectual property industry, providing specific steps that every business leader can take to implement them.
Understand the limitations
ChatGPT is a powerful tool, but it is not a legal expert. While it can help generate ideas and suggestions, it should not be relied upon for making legal decisions. Think of it more like a hyper-caffeinated intern. It will excitedly return results, quickly, but don’t trust it. Check and double check ChatGPT’s work. Make sure it’s not only what you would say, but how you would say it. Context and cadence matter. Intellectual property strategy can be made or broken by a few choice words, knowing who needs what information, when. To mitigate the risks associated with relying too heavily on the output generated by ChatGPT, it’s important for businesses to:
- Assign a technical or legal expert to review and interpret any output generated by ChatGPT. This expert can ensure that any recommendations are in line with business requirements and ethical standards.
- Invest in training programs that provide employees with a clear understanding of the limits of ChatGPT. This will ensure that employees are not relying on ChatGPT for legal decision-making.
Example:
A manager at a tech startup could assign engineering and legal experts to review and interpret patent claims generated by ChatGPT. This expert can ensure that any claims generated by the tool actually cover the broadest version of the technology. Let ChatGPT create the rough draft, not the final version.
Protect confidential information
Language models, like ChatGPT requires input to generate output. However, businesses must ensure that they are not inputting or generating confidential information that could be used to infringe on the intellectual property rights of others. Exposing trade secret information to ChatGPT creates a real business risk. Regardless of whether or not you trust OpenAI with your confidential information, you should note that it’s a rabidly developing technology platform that is susceptible to theft and malfunction. OpenAI recently apologized for leaking payment information and threads to other users. To protect confidential information, businesses can:
- Develop protocols for handling confidential information. These protocols should include clear guidelines for how to handle sensitive data when using ChatGPT.
- Restrict access to ChatGPT to only those employees who need to use it for business purposes. This can help ensure that confidential information is not leaked.
Example:
A business leader could restrict access to ChatGPT to only those employees who have been trained in handling confidential information. This can help ensure that confidential information is not leaked when generating invention disclosures, legal briefs, quarterly/annual reports, or other documents.
Verify information
It bears repeating: ChatGPT doesn’t “know” the “truth” about any subject. But it is very good at sounding like its answers are fact-checked and evidence based. It’s important to double-check any information generated by ChatGPT to ensure that it is accurate and up-to-date. To verify information, businesses can:
- Cross-check output generated by ChatGPT against other sources of information. This can help ensure that the output is accurate and relevant.
- Develop a quality control process that includes reviewing any output generated by ChatGPT. This can help catch any errors or inaccuracies before they are acted upon.
Example:
An industrial business leader could cross-check output generated by ChatGPT against technical manuals or other sources of information. This can help ensure that any recommendations generated by the tool are accurate and reliable.
Use clear language
To ensure that ChatGPT generates accurate and relevant output, it’s important to use clear and concise language when communicating with the tool. To use clear language, businesses can:
- Develop a glossary of industry-specific terms and acronyms. This can help ensure that employees are using language that ChatGPT can understand.
- Provide employees with training on how to communicate effectively with ChatGPT. This can include tips on how to phrase questions and how to use punctuation to ensure that the output generated is clear and concise.
Example:
A high-tech company could develop a glossary of industry-specific terms and acronyms that employees can use when communicating with ChatGPT. This can help ensure that the tool generates accurate and relevant output when conducting patent searches or other research. You can start a thread by dumping a boilerplate explanation of technical terms to ChatGPT so you are speaking the same language.
Consider context
When using ChatGPT, it’s important to understand the context in which the tool is being used and the potential implications of any output generated. When it comes to intellectual property, context is king. There are elements of documentation and communication that are tedious and low-value-add, this is where ChatGPT can speed things up and outperform a human. However, there are many scenarios where a human-touch is required where ChatGPT will at best be an awkward approach, at worst completely bungle the job. To consider context, businesses can:
- Review the output generated by ChatGPT in the context of the business goals and objectives. This can help ensure that any recommendations generated by the tool are aligned with the overall strategy.
- Develop a risk management plan that includes a review of any potential legal or ethical issues associated with using ChatGPT. This can help ensure that any risks associated with the tool are identified and addressed proactively.
Example:
A law firm could review the output generated by ChatGPT in the context of the legal case or matter at hand. This can help ensure that any recommendations generated by the tool are aligned with the overall legal strategy.
Provide feedback
Providing feedback to ChatGPT is essential for the tool to continue learning and improving its output. Remember, you can’t hurt ChatGPT’s feelings. Tell it when it missed the mark and how it can improve, and you may be surprised how well it can course correct. To provide feedback, businesses can:
- Develop a process for employees to provide feedback on the output generated by ChatGPT. Keep a list of prompts hand that can help coach ChatGPT to move in the right direction. If you see a result going off the rails, stop generating and reassess the original prompt.
- Encourage employees to have conversations about how to use ChatGPT. Keep a thread open in your preferred workspace where employees and managers can share tips and tricks about how to get the results you want out of ChatGPT
Example:
In an industrial setting, a business leader could encourage employees to provide feedback on the output generated by ChatGPT when conducting maintenance or repair task documentation. Keep a running list of what’s work well and where ChatGPT seems stumped. This can help ensure that the your interaction with the tool continues to improve over time.
Comply with ethical standards
Finally, it’s important to use ChatGPT in a way that is consistent with ethical standards and legal requirements. I saved the best for last because while business ethics should always be at the forefront of discussions with new technology, it’s often easy to forget when an exciting new tool like ChatGPT is released into the wild. To comply with ethical standards, businesses can:
- Develop a code of conduct that includes guidelines for using ChatGPT. This can include guidelines for handling confidential information, communicating effectively with the tool, and ensuring that the tool is used in a way that is consistent with legal and ethical standards.
- Review and update the code of conduct regularly to ensure that it remains current with changing legal and ethical standards.
Example:
A business leader could develop a code of conduct that includes guidelines for using ChatGPT when conducting research and development activities. This can help ensure that the tool is used in a way that is consistent with legal and ethical standards for intellectual property protection.
Remember that ChatGPT is a completely new paradigm for knowledge work. Generative AI is here to stay, and anyone who claims otherwise will soon sound like skeptics of the power of personal computers or the internet, luddites of all stripes.
Businesses in the intellectual property industry can benefit from using ChatGPT as a tool to assist with legal research and analysis. However, it’s important to use the tool responsibly and in a way that is consistent with legal and ethical standards. By following these best practices and implementing the specific steps outlined above, businesses can maximize the benefits of ChatGPT while minimizing the risks associated with its use.