Contributed by Carlee Mixon, Space Coast Chapter
Ready or not, the realm of artificial intelligence (AI) is hurtling forward at a rapid rate, and failing to keep pace could lead us down a path of obsolescence. The exciting reality is that AI is poised to completely transform our work processes and communication dynamics, unlocking extraordinary possibilities. However, the flip side of this coin is the uncertainty surrounding the exact nature of this transformation. With new AI product announcements from tech giants like Microsoft, Google, Apple, Open AI, Meta, LinkedIn and Adobe, the opportunity arises for us to embrace AI as a catalyst for enhancing workforce productivity rather than succumbing to apprehension.
During the FPRA Annual Conference General Session, “AI, Ethics and Reputation: An Existential Challenge for PR,” Martin Waxman, MCM, APR, president of Martin Waxman Communications and instructor at LinkedIn Learning, explored the ethical challenges around AI and PR, including data security, privacy, safety, bias and reputation.
3 Types of AI
- Narrow: Trained on lots of data. Makes predictions on what it knows.
- General: Learns on the fly. What all the big tech companies are spending billions of dollars on trying to create.
- Super: When the intelligence eclipses the intelligence of humanity
Key Ethical Issues
- Privacy & Safety
- Relational AI
Undoubtedly, bias has emerged as the foremost concern within the realm of AI today. Even when data originates from seemingly reputable sources, it’s imperative that we extend extra efforts to verify its accuracy. Bias infiltrates AI through various avenues: from the data itself to the algorithms, developers, and even the cognitive processes involved. These diverse forms of bias hold far-reaching implications, influencing our understanding, recommendations, content delivery and the very foundation of trust in AI systems. The pivotal role of data cannot be overstated, as it encompasses not only textual information but also the very fabric of our spoken words and visual content.
Privacy and Safety
In embracing the capabilities of AI, we must be diligent in upholding these principles to create a safer, more respectful and trustworthy technological landscape.
- Data sources, collection, storage, permissions
- Exploitation of personal identifiable information (PII)
- Facial recognition
- Data breaches
- Racist and harmful language
Is your data safe? Can it be breached? Waxman mentioned that Zoom recently updated their terms of service to include that they are using calls and meetings to learn.
Explainability in AI stands as a crucial bridge between complex deep learning algorithms and human comprehension. In an age where AI influences pivotal decisions across domains such as health, finance, education and beyond, the demand for understanding the ‘why’ behind key algorithmic choices is paramount for ensuring fairness and accountability. This is especially vital in sensitive areas that deeply impact lives. Notably, industry leaders like IBM are taking strides to tackle this challenge by developing an explainable AI toolkit, facilitating the interpretation of AI-driven outcomes. Transparency and disclosure emerge as guiding principles, empowering users to navigate the AI landscape with clarity and confidence. Through these efforts, the trajectory of AI is not just about efficiency, but also about building a foundation of trust and comprehension in the evolving technological landscape.
The world of AI brings forth a multitude of intriguing questions surrounding ownership. Who possesses the rights to the distinct sound of your voice, or the outputs generated by your AI assistant? With innovations like Microsoft’s Vall-E, capable of synthesizing voices from mere 3-second clips, the notion of ownership becomes intricate. Delving into the realm of creativity, an equally compelling query emerges: if an AI output mirrors your artistic style or tone, how does compensation play out? Navigating the landscape of AI ownership prompts us to grapple with not only the technical intricacies but also the fundamental rights and responsibilities inherent in this evolving domain.
Relational AI is forging a new era of human-AI interactions, with chatbots taking on roles as innovative intermediaries. This evolution is exemplified by platforms like Chat.D-ID, seamlessly integrating voice communication with responsive synthetic avatars, blurring the line between human and AI interactions. CarynAI, launched by a Snapchat influencer, introduces the concept of a ‘virtual girlfriend,’ underscoring the transformative potential of AI in relationships. This shift redefines power dynamics, with machine-driven persuasion becoming increasingly relevant. Chatbots, meticulously designed to be both sensible and specific, are ushering in a wave of meaningful conversations, giving rise to novel forms of engagement and connection in the digital age.
Develop a Principles-Base Approach to AI Ethics
- LEARN about AI data
- DEFINE the PR and AI pitfalls
- IDENTIFY ethical issues and PR principles
- USE decision making tree
- DECIDE ethically
Adopt a Strategic Approach
- Consider AI impact on audiences. How is this going to affect relationships? Internal and external?
- Integrate ethics from development to implementation of AI systems
- Manage risks and reputation
- Re-map creative workflow
- Imagine wide-ranging consequences. Crisis planning – think about science fiction scenarios
- Secure and protect your data. What are you doing to make sure your data is clean and safe?
- Ensure a human has the final say