THE MORAL IMPERATIVES OF GENERATIVE AI IN EDUCATION
The swift progress of AI is transforming the educational domain, sparking vital dialogues regarding its ethical implementation. The University of Manchester has introduced a set of guidelines to provide clear direction on the judicious assimilation of potent generative tools like ChatGPT within educational contexts.
This structure promotes trial and underscores the enhancement of human capabilities while acknowledging possible issues such as inaccuracy and bias. As AI models advance, this code of conduct aims to facilitate the responsible development of education by harmonizing the strengths of both humans and AI, underpinned by principles of academic integrity, transparency, and purposeful adoption.
According to the AI Code of Conduct, we are able to look at ethical implications of using AI in contexts such as everyday use, rules for learning assessments, and an ethical framework for disseminating outputs. It argues these tools should enhance human skills rather than replace them entirely.
The Code permits using Gen AI to aid idea generation and efficiency but warns us against academic malpractice through fabricated content. These malpractices can include citing references and conducting other forms of academic dishonesty. With that being said, clear rules are outlined for referencing and attributing any AI-generated text, images or data used in submissions. This includes quoting content, citing the tool and version, annotating inaccuracies, and critically evaluating relevance.
The study reveals most students already utilize these tools to brainstorm concepts, accelerate writing, and simplify research. However, both students and staff want ethical oversight and training to maximize benefits while minimizing potential downsides.
Several limitations of current AI models are analyzed including inaccuracy due to biased datasets, lack of contextual awareness, tendency to fabricate facts and sources, and the inability to provide in-depth analysis. The code thus stresses that human judgment remains essential for verifying quality and factual correctness.
Issues with Accuracy and Reliability
AI generative tools can produce content that lacks accuracy or reliability due to incomplete or biased training data. Errors, nonsensical content, and contradictions are possible. The quality of training data impacts the accuracy of generated content.
Knowledge Limitations
AI generative tools have data limitations, meaning they only know information up to a certain date. This can lead to outdated or incomplete responses when asked about recent events. For example, ChatGPT (GPT3.5) relies on data up to September 2021.
Non-Deterministic AI Models
AI generative models can give varied responses for the same input. If you ask the same question repeatedly, you might get different answers. This unpredictability can be a limitation, especially in applications where consistency is crucial.
Perpetuating Bias in AI Models
AI models learn from their training data. If this data contains biases or stereotypes, the models may reproduce those biases in their generated content. This can reinforce societal prejudices and exclude diverse viewpoints.
Misinformation
AI generative tools can generate inaccurate information because they attempt to anticipate the next word or word sequence in a given sentence. This tendency can lead to the creation of false facts or information.
Lack of Diversity in Cultural
AI generative tools might show a preference for western viewpoints because they are primarily trained on data from western sources. This can result in underrepresentation and limited understanding of other cultures and perspectives.
Sign up to enjoy an exclusive early-access invitation and be a pioneer in transforming digital education!
GOOD OVERT GEN AI
From this, we can learn that ethical framework calls for protecting user autonomy by validating outputs, ensuring transparency on methods, fostering accountability for material, and promoting inclusivity by considering impacts on diverse groups. This applies when disseminating AI content publicly.
We are also able to learn that the education sector is encouraged to advocate experimenting with generative AI. However, this will have to include maintaining academic integrity through human oversight. If used conscientiously, AI can expand creativity, personalization and access. However, verification, ethics and academic principles are crucial to minimize risks and biases. Education must evolve alongside technology responsibly.
AI Code of Ethics is organised into three levels, as outlined in the image below.
Tier 1: Guidance for Personal Use of Generative AI Tool
Tier 1 of the AI Code of Ethics focuses on individual use of generative AI tools. It emphasizes several key guidelines for users to follow when engaging with these tools. Users are advised to be aware of the AI functionality embedded within certain tools and to refrain from sharing sensitive information in input prompts, as such data may contribute to model training. Additionally, it cautions users against relying on AI-generated content as advice in areas like legal, health, or financial matters
Tier 2: Guidelines for Incorporating Generative AI in Education and Assessment
The academic team will review submissions to identify instances of AI-generated content misuse, as outlined in the following criteria. If any inappropriate use is detected, the submission will be subject to further scrutiny for potential academic misconduct. In such instances, grades will be withheld, and students will undergo an investigation for academic misconduct.
Tier 3: Ethical Framework for AI Application
The application of generative AI tools must consistently align with ethical principles defined below. These ethical standards have been collaboratively developed with students participating in the AI: Robot Overlord UCIL module, underscoring the imperative for ethical and transparent utilisation of such tools.
What are you waiting for?
Sign up to enjoy an exclusive early-access invitation and be a pioneer in transforming digital education!
OTHER PROJECTS
Browse our other exciting GEN AI innovation projects and see how we are pushing the boundaries beyond what's possible with GEN AI.
COMING SOON !
We can't wait to launch it and we believe you are going to love our innovations. Stay tuned.
