Generative AI for Social Work Students: Part II
Essential knowledge, abilities, and practices with AI
This article is the second in a series introducing social work students to generative AI. I recommend starting with the first article, which you can find here:
Mastering AI tools like ChatGPT offers MSW students a range of career advantages. Proficiency in these technologies streamlines work processes, increasing efficiency and accuracy in tasks like case management, data analysis, and resource allocation. As technology becomes increasingly integrated into social work practice, organizations will value professionals who seamlessly blend traditional skills with AI’s innovative capabilities. Proficiency in AI tools demonstrates a commitment to staying current with technological advancements, making students more marketable to potential employers.
The ability of AI to revolutionize social work practice depends on professionals having a solid grasp of the technology, which includes understanding how AI models function, their strengths and weaknesses, and the potential issues and biases they may introduce. Even though ChatGPT has only been publicly accessible for under six months, I have observed problematic usage among my students. Some of these issues stem from overconfidence in their abilities while lacking the required skill to effectively use these powerful tools, a phenomenon known as the Dunning-Kruger effect.
As the social work curriculum has yet to incorporate this technology fully, this article serves as a stop-gap measure, offering guidance on essential knowledge, skills, and practices for working with generative AI. While not a comprehensive roadmap, this article provides a starting point that will evolve with the technology. By refining their abilities in AI applications like ChatGPT, social work students can contribute to developing novel, cutting-edge interventions and practices, positioning themselves as thought leaders. This expertise can open doors for career growth, professional recognition, and collaboration with multidisciplinary teams to effect lasting, positive change for individuals and communities.
Given the swift advancements in AI technologies and abundant available resources, I refrain from offering a specific reading list, as it would quickly become outdated. Instead, I present these areas as general competency domains, which you should consistently monitor for the most up-to-date resources.
Understanding how artificial intelligence works
In today’s fast-paced technological environment, social work students must develop an understanding of artificial intelligence (AI), especially large language models like ChatGPT. As AI tools become more prevalent in social work practice, professionals with expertise in this area can make well-informed decisions about their usage. This knowledge helps social workers use AI responsibly and ethically while integrating their content expertise and interpersonal skills.
A key component of understanding AI is knowing the underlying training procedures and data, which can help you make informed decisions about when and how to use AI-generated content. This knowledge also enables you to identify potential inaccuracies, biases, and outdated information. Evaluating an AI model’s performance for specific use cases is essential. Disappointment with AI models often arises from using them for tasks they were not trained to perform, such as researching highly specialized topics.
It is important to note that understanding AI and large language models does not require a background in math, statistics, or computer programming. While achieving proficiency involves time and effort, it is an attainable goal for any social work student committed to learning. A wealth of resources, including online courses, tutorials, and articles, is available to help students develop their AI knowledge at their own pace without requiring specialized prerequisites.
By investing in this learning process, social work students can enhance their skill sets and contribute to advancing the field. In doing so, they ensure that they harness the power of AI to benefit individuals, families, and communities in a way that aligns with the principles of social work.
Ethical issues
The significance of actively engaging with and staying informed about ethical issues related to artificial intelligence cannot be overstated. Ensuring the responsible use of AI, especially large language models like ChatGPT, is critical to aligning the technology with the core values and principles of social work practice, such as social justice, the dignity and worth of individuals, and the significance of human relationships.
Ethical concerns in AI use cover a wide range of issues, including:
- Data privacy: Safeguarding the sensitive information of clients and communities is essential to maintaining trust and upholding ethical standards in social work practice.
- Biases in AI systems: AI models can inadvertently perpetuate and reinforce existing biases, potentially leading to discriminatory outcomes. Social workers must be aware of these biases and actively work to mitigate their impact.
- Potential misuse: Unintended or unethical use of AI tools can harm individuals and communities, making it vital for social workers to understand and adhere to appropriate guidelines for using such technology.
- Equitable distribution of AI benefits: Ensuring that the advantages of AI technology are accessible to all, regardless of socioeconomic status or other factors, is critical to fostering social justice and addressing systemic inequalities.
As social workers, our responsibility is to continuously examine and question the effects of AI technology on vulnerable populations, ensuring that the tools we use do not contribute to harm or injustice. Cultivating a deep understanding of these ethical challenges enables social work professionals to advocate for developing and implementing policies and guidelines that support ethical AI use in their practice settings. This commitment to ethical AI integration is essential for upholding the core values of social work and promoting the well-being of individuals and communities.
Prompt engineering
Currently, “prompt engineering” is one of the most important practical skills to learn to harness the power of generative AI tools. Prompt engineering may sound complicated, but the concept is straightforward. In the context of generative AI and large language models, prompt engineering refers to crafting practical questions or input statements (i.e., prompts) to guide the AI model in generating useful, relevant, and accurate responses.
Learning to write effective prompts is essential for using generative AI tools to their full potential. Well-crafted prompts can significantly improve the quality of the AI-generated content, making it more focused, accurate, and applicable to the task at hand. Conversely, poorly designed prompts may lead to irrelevant, ambiguous, or misleading responses from the AI model.
To engineer effective prompts, one must deeply understand the subject matter and anticipate potential pitfalls and inaccuracies that may arise in AI-generated content. In this regard, content expertise plays a significant role. Professionals with solid knowledge in a content area can design prompts that will lead to meaningful and insightful AI-generated content more effectively.
The significance of subject matter proficiency in designing AI prompts emphasizes that generative AI cannot substitute for social workers. AI models depend on human experts’ insights and experiences to shape their responses and guarantee the produced content is pertinent and suitable for a particular context. Social workers’ expertise enables them to efficiently utilize AI tools to improve their work while ensuring the AI-generated content adheres to the profession’s values, principles, and ethical guidelines.
As generative AI continues to play a more prominent role in various fields, including social work, prompt engineering is emerging as one of the essential AI-related skills for professionals to learn. By mastering prompt engineering, social workers can unlock the full potential of AI tools, enhancing their practice and driving more effective, efficient, and ethical outcomes for their clients and communities.
Design thinking
Design thinking is an approach that emphasizes a human-centered methodology for addressing complex problems. This approach emphasizes empathy, collaboration, and experimentation to understand the unique needs and contexts of the communities involved. We can create more effective and tailored AI-driven solutions to address intricate social issues by incorporating design thinking principles into generative AI tools.
This fusion of design thinking and generative AI fosters a creative environment that encourages idea generation and exploration of diverse solutions. It also supports an iterative process of prototyping and testing, allowing for the continuous refinement and adjustment of interventions based on real-world feedback. Adopting a design thinking mindset when utilizing generative AI tools facilitates interdisciplinary collaboration and promotes a culture of continuous improvement. This enables the development of comprehensive solutions to complex social issues by drawing on a wide range of perspectives and expertise from different fields.
Moreover, by learning from failures and iterating on solutions, we can ensure that AI-driven interventions are consistently refined and adapted to meet the changing needs of the target population. This approach ultimately leads to more effective, responsive, and sustainable solutions that address the root causes of social issues and improve the well-being of the communities served.
Next steps
The following two articles in this series will focus on strategies for enhancing your educational experience with AI and a concluding article on the art and science of prompt engineering. As I prepare these articles, I’m also actively examining the problems of the ASWB national social work licensing exam using large language models. This work is highly relevant to this series and the field of social work. You can find my most recent article on this topic here:
