Scribble at 2023-06-02 11:52:33 Last modified: unmodified

Prompt: People talk about risks to use ChatGPT in business and daily life.

Response: Yes, that's correct! There are concerns about the potential misuse of AI models like ChatGPT, especially when it comes to sensitive information or decision-making processes. It is important for organizations to establish guidelines and best practices for using these tools responsibly.

Prompt: What do you think of "Statement on AI Risk" at https://www.safe.ai/statement-on-ai-risk ?

Response: The "Statement on AI Risk" is a document that outlines the potential risks and challenges associated with artificial intelligence, including issues related to bias, security, and ethics. It provides a framework for organizations to consider when developing and deploying AI systems, and emphasizes the importance of responsible development and use of these technologies. Overall, it is an important resource for anyone working with AI, as it helps to ensure that these tools are used in a way that benefits society while minimizing potential harms.

Prompt: Yes. How do you know what are risks in usage of AI?

Response: There are several ways to identify and understand the risks associated with using AI. One approach is to conduct a risk assessment, which involves analyzing the potential impacts and consequences of an AI system on individuals, organizations, and society as a whole. This can involve gathering information about the specific use case, identifying any potential biases or vulnerabilities in the model, and considering how the system might be used to make decisions that could have significant consequences. Additionally, it is important for organizations to stay informed about new developments and advancements in AI technology, as well as any emerging risks or challenges associated with these tools. By taking a proactive approach to understanding and addressing the potential risks of using AI, organizations can help ensure that these technologies are used responsibly and effectively.

Prompt: Is it just like an old phrase of "Privacy by Design" coined by a privacy commissioner of Canada to find an estimate risks in usage of system to manage PII of customers?

Response: Yes, that's correct! The concept of "privacy by design" involves incorporating privacy considerations into the development and deployment of systems from the outset, rather than as an afterthought. This can help organizations identify potential risks associated with handling sensitive information, such as personal data, and take steps to mitigate those risks before they become a problem. By considering privacy and security concerns early in the process, organizations can create more secure and effective systems that are better equipped to handle sensitive information while minimizing the risk of harm or misuse.

Prompt: Yes. So, we have taken a policy and controls in business so that anyone should not easily trust the result of ChatGPT in search of personality of job applicant and employee.

Response: That's correct! By implementing policies and controls around the use of AI systems like ChatGPT, organizations can help ensure that these tools are used responsibly and effectively. This may involve establishing guidelines for how to handle sensitive information, such as personal data, or setting up safeguards to prevent unauthorized access or misuse of the system. By taking a proactive approach to managing risks associated with AI technology, organizations can help ensure that these tools are used in a way that benefits society while minimizing potential harms.

以上は、gpt4all でのやりとりだ。最近はオンラインの ChatGPT ではなく、もっぱらローカル環境で使っているのだが、こちらの方がネットワークを使っていないだけ、レスポンスが速い場合がある。ちなみに、Microsoft の Editor という校正ツールが入っていて、こうしてフォームに入力すると自動で添削してくれるのだが、僕の文章は添削箇所がゼロなのに、ChatGPT の文章には添削箇所が幾つかある。これは或る意味では AI vs. AI なのだが、僕が見ている限りでは ChatGPT の英語のセンスの方が正しいと思う。