Over the last little while I have had many people asking about AI and education (I live in a place that is also a native habitat of university lecturers and educators). The key concerns expressed are that students will instantly become serial cheaters and use ChatGPT to write their essays and that they will lose their ability to think critically. And then there are all the other standard moral panics that accompany change. This indicates to me that for all the recent hype most of these people have had limited access to, or active experience of, using AI. Although AI might be present in customer bots and some apps they have never consciously set out to use it. If they had they would learn something pretty quickly — AI is really, really useful.
ChatGPT, for example, is not just a chat bot. The underlying models when coupled with an emerging array of add-ons or plugins such as the Advanced Data Analysis plugin, allows ChatGPT to quickly perform data analysis. To do this you can upload a spreadsheet and it will not only return a summary of the data the spreadsheet contains, but will also generate charts to include in your reports. In the world of film making, AI is also contributing enormously. Generative AI is used to repair and reconstruct damaged footage, to allow advanced animation of characters (my digital collaborator Fin being an example), in motion control systems for cameras, virtual production and so on. This is still only scratching the surface.
These are all examples of what is termed ‘weak’ AI or ‘narrow’ AI. Weak AI refers to artificial intelligence systems that are designed to perform specific tasks or solve specific problems without possessing consciousness, self-awareness, or genuine understanding. We are a long way off the ‘conscious’ AI that will decide to wipe us off the planet. The risk of human annihilation by AI exists simply because someone has programmed it that way, not because it has the capability to decide that we were annoying insects.
Like fire or electricity, what we call AI is amazingly powerful, useful and, if not handled correctly, harmful. It is already driving innovation and effecting social change. So the question I would put to my university buds when they come to me with their concerns about AI is not what harms it might cause to education, but what harms might arise from not educating students about it.
I threw the following prompt at ChatGPT about the harms of not teaching (film) students about AI and it gave me a neat, bulleted list:
ChatGPT, for example, is not just a chat bot. The underlying models when coupled with an emerging array of add-ons or plugins such as the Advanced Data Analysis plugin, allows ChatGPT to quickly perform data analysis. To do this you can upload a spreadsheet and it will not only return a summary of the data the spreadsheet contains, but will also generate charts to include in your reports. In the world of film making, AI is also contributing enormously. Generative AI is used to repair and reconstruct damaged footage, to allow advanced animation of characters (my digital collaborator Fin being an example), in motion control systems for cameras, virtual production and so on. This is still only scratching the surface.
These are all examples of what is termed ‘weak’ AI or ‘narrow’ AI. Weak AI refers to artificial intelligence systems that are designed to perform specific tasks or solve specific problems without possessing consciousness, self-awareness, or genuine understanding. We are a long way off the ‘conscious’ AI that will decide to wipe us off the planet. The risk of human annihilation by AI exists simply because someone has programmed it that way, not because it has the capability to decide that we were annoying insects.
Like fire or electricity, what we call AI is amazingly powerful, useful and, if not handled correctly, harmful. It is already driving innovation and effecting social change. So the question I would put to my university buds when they come to me with their concerns about AI is not what harms it might cause to education, but what harms might arise from not educating students about it.
I threw the following prompt at ChatGPT about the harms of not teaching (film) students about AI and it gave me a neat, bulleted list:
- If universities, lecturers, and students in film making do not engage with AI and applications like mine, they face several significant threats:
- Falling Behind in Industry Standards: The film industry is rapidly integrating AI for various purposes, from script analysis to post-production. Not engaging with AI means students and educators might lack familiarity with these emerging tools, putting them at a disadvantage in the job market.
- Limited Creative Possibilities: AI offers new avenues for creativity, like generating unique visual effects or aiding in complex story development. Ignoring these tools may limit the creative scope and innovation potential for students and lecturers.
- Inefficient Processes: AI can streamline many time-consuming tasks, such as editing or color grading. Without these efficiencies, film students and educators might spend more time on technical tasks than on creative ones, reducing overall productivity.
- Narrow Skill Set: As the industry evolves, a broad skill set that includes AI competency is becoming more valuable. Not learning these skills could make students less competitive compared to peers who are proficient in AI.
- Limited Analytical Insights: AI can provide valuable insights through data analysis, like audience preferences and trends. Not utilizing AI means missing out on these insights, which could guide more informed decisions in filmmaking.
- Reduced Adaptability: The ability to adapt to new technologies is crucial in the dynamic film industry. A reluctance to engage with AI and similar applications could hinder adaptability, making it harder for individuals to stay relevant as the industry evolves.
To get this list, however, I had to write an effective prompt. A prompt is a written or spoken request made to the AI model to elicit information. It follows the old dictum of crap-in-crap-out. There is a n emerging discipline in prompt writing and even prompt engineering in software development (OpenAI, 2023) because writing an effective prompt is not easy. In fact, it is an exercise in logic, clear thinking, and problem solving.
Here is ChatGPT’s checklist for writing an effective prompt:
If only our students could do this…
References
Marr, B. (2023). The 15 Biggest Risks Of Artificial Intelligence [online]. Available from: https://www.forbes.com/sites/bernardmarr/2023/06/02/the-15-biggest-risks-of-artificial-intelligence/ [Accessed 17 November 2023].
OpenAI (2023). Best practices for prompt engineering with OpenAI API | OpenAI Help Center [online]. Available from: https://help.openai.com/en/articles/6654000-best-practices-for-prompt-engineering-with-openai-api [Accessed 17 November 2023].
- Be Specific: Clearly state what you need. For example, instead of asking “Tell me about history,” specify “Please provide an overview of the French Revolution.”
- · Provide Context: If your question builds on previous information or requires specific knowledge, include that context. For instance, “Following our discussion on renewable energy, how do solar panels impact the environment?”
- Use Clear and Direct Language: Avoid ambiguity and overly complex sentences. Direct and simple language helps in getting precise answers.
- State Your Purpose: If you have a specific goal, like needing information for a project or understanding a concept, mention it. For example, “I need detailed information on photosynthesis for my biology report.”
- Indicate Desired Detail or Length: If you have a preference for the length or depth of the response, state it upfront. For example, “Can you provide a brief summary of World War II?”
- Include Necessary Constraints: If there are any constraints such as word count, specific formats (like APA or MLA for citations), or certain viewpoints, mention them.
- Ask One Question at a Time: This ensures clarity and makes it easier to get a direct answer to each specific query.· Avoid Jargon Unless Necessary: Use terms that are easy to understand unless your question requires technical language.
- Be Open to Follow-Up Questions: Sometimes, your prompt may need clarification, so be prepared to provide additional information if asked.
- Review Before Sending: Quickly review your prompt to check for clarity and completeness.
If only our students could do this…
References
Marr, B. (2023). The 15 Biggest Risks Of Artificial Intelligence [online]. Available from: https://www.forbes.com/sites/bernardmarr/2023/06/02/the-15-biggest-risks-of-artificial-intelligence/ [Accessed 17 November 2023].
OpenAI (2023). Best practices for prompt engineering with OpenAI API | OpenAI Help Center [online]. Available from: https://help.openai.com/en/articles/6654000-best-practices-for-prompt-engineering-with-openai-api [Accessed 17 November 2023].
Created with
