While I was learning data analytics last year through Merit America's Data Analytics program, I was also first introduced to prompt engineering and generative AI by learning how to use ChatGPT. We used this tool mostly for job search preparation, such as generating a job resume and cover letter. We send prompts to ChatGPT based on our existing sources, such as the desired job opening's job description and our personal skills, experience, and educational background. While it may generate strong and effective resumes and cover letters, we mainly use them as starter references and we still have to craft our resumes and cover letters with our signatures and touches. ChatGPT was a marvel to me and somehow has given me some kind of ability to craft original pieces.
If you want to be truly authentic, you do not copy and paste.
That being said, ChatGPT and generative AI have become very useful tools that can make our daily tasks easier and faster. However, I am also an artist, and I take the authenticity of my artwork very seriously. When I see an AI-generated digital artwork with so many inconsistencies and similarities to already existing artwork floating around social media, it's very upsetting. If you don't follow the art community on the internet, there is a very loud outcry from the artists against AI-generated art.
When I used to have my old account on deviantART over 20 years and deviantART began supporting AI art, the majority of the new artwork I've seen in their main gallery has been AI-generated. I dabbled with AI-generated art before out of curiosity, and while the result was quite impressive, I did not feel satisfied. There were strange inconsistencies and oddities, and in addition, I didn't feel that the finished result had any artistic soul in it. It was completely digitalized plagiarism all in one art piece.
With that being said, I'm very impressed with Generative AI altogether, only if it helps the most important fields, such as job search, healthcare, education, public services, and more. However, at the same time, Generative AI should also have its limits and should not be used for unethical, dishonest purposes, such as generating AI artwork for dishonest authenticity or manipulating existing photos for malicious purposes, such as with tabloids or causing harm to others.
This is where this thing called Ethical AI comes in.
What is Ethical AI?
I asked ChatGPT about Ethical AI, and this is what ChatGPT gave me:
Ethical AI refers to the development and deployment of artificial intelligence systems in a manner that is fair, transparent, responsible, and aligned with moral and societal values. The concept of ethical AI involves ensuring that AI technologies do not cause harm, reinforce biases, or perpetuate inequality, and that they are used to benefit individuals and society as a whole. Key aspects of ethical AI include:
1. Fairness and Non-Discrimination
Avoiding Bias: Ethical AI systems should be designed and trained in a way that avoids biases based on race, gender, age, or other protected characteristics. This involves careful consideration of the data used to train AI models and the potential for biased outcomes.
Equitable Access: AI technologies should be accessible to diverse populations and not exacerbate existing inequalities.
2. Transparency and Explainability
Clarity on Decision-Making: AI systems should be transparent in how they make decisions. Users and stakeholders should be able to understand the factors that influence the AI's outputs.
Explainability: The rationale behind AI decisions should be explainable to non-experts, ensuring that users can trust and validate the system's actions.
3. Privacy and Data Protection
Data Privacy: Ethical AI must respect the privacy of individuals by securely handling personal data and ensuring that data collection and usage are done with informed consent.
Data Security: AI systems should have robust security measures to protect data from breaches or unauthorized access.
4. Accountability and Responsibility
Clear Responsibility: Organizations developing and deploying AI systems should take responsibility for the actions and outcomes of those systems. This includes establishing clear accountability for errors, biases, or harm caused by AI.
Regulatory Compliance: Ethical AI must comply with laws, regulations, and guidelines governing its use, including those related to data protection, discrimination, and human rights.
5. Human-Centric Design
User Empowerment: AI systems should be designed to augment human capabilities rather than replace or undermine human decision-making. Users should have the ability to override or question AI decisions.
Well-being: AI should be developed with the well-being of individuals and society in mind, avoiding actions that could cause harm or negative societal impact.
6. Sustainability and Environmental Impact
Resource Efficiency: Ethical AI should consider the environmental impact of AI technologies, including the energy consumption and carbon footprint associated with training and deploying AI models.
Sustainable Practices: AI development should align with broader goals of sustainability and contribute positively to environmental protection.
7. Inclusive Design and Stakeholder Involvement
Broad Participation: The design and deployment of AI systems should involve a wide range of stakeholders, including those from marginalized or underrepresented groups, to ensure that diverse perspectives are considered.
Inclusive Benefits: AI should be used to benefit all parts of society, including vulnerable and underserved communities.
8. Ethical Use Cases
- Beneficial Applications: AI should be applied in ways that contribute positively to society, such as improving healthcare, education, and public services, rather than in ways that could cause harm, such as in autonomous weapons or invasive surveillance.
Ethical AI is about creating AI systems that align with human values and societal norms, ensuring that the technology serves humanity in a just, equitable, and responsible manner. It requires ongoing dialogue, interdisciplinary collaboration, and a commitment to ethical principles throughout the AI lifecycle.
There are many articles about artists against generative AI, such as this one from Forbes. I encourage everyone to search around the internet and read these news articles. I don't feel the need to repeat the reasons as these articles have already voiced how I feel about recreational generative AI.
The ChatGPT result of Ethical AI I posted above may not be complete or even accurate. However, Point #8 (Ethical Use Cases) should cover AI-generated art.
I do like AI in general, to be honest. I feel that AI can be a valuable tool for data analytics, data science, and machine learning (and pretty much every subject out there). However, I am against recreational generative AI, where you may think you have created something original through generative AI, only to have been mashed up from elements from existing artwork. I'm still a newbie when it comes to AI, as I don't use ChatGPT frequently myself.
Once I learn more about Ethical AI, I hope to write more about it here in the near future.