Generative AI (GenAi), a rapidly evolving technology, holds promise for amplifying public sector efficacy, particularly in the fields of public health preparedness and emergency management. While technical and creative capabilities are paramount for leveraging AI, equally crucial are the critical thinking skills to question and ensure its responsible use.
But the question remains: are you AI ready?
This piece will help you by delineating:
The essence of generative AI
Its operational mechanics
Requisite skills for mastery
Critical questions to ask before and while using AI
What is Generative AI
Generative AI encompasses artificial intelligence systems optimized for crafting new content—text, images, audio, video. Unlike other AI, like expert systems or computer vision, generative models make original outputs, sometimes indistinguishable from human-created content. ChatGPT, Claude, and Bard are a few examples of Generative AI, but there many other applications.
Within the public sector, this could revolutionize information analysis and dissemination, situational awareness, and crisis communication.
Mechanics of Generative AI
Employing vast datasets, generative AI deciphers patterns and relationships among data points. For instance, text generation systems like GPT-4 are tutored on volumes of text, learning language structures and semantics. By employing deep learning algorithms and neural networks, these models discern how to fabricate realistic outputs based on the patterns recognized.
GenAI (like ChatGPT) training data may not have undergone fact-checking, enabling the model to fabricate responses to users' inquiries even in the absence of factual information -- a scenario referred to as hallucination. The uncurated nature of the data often embeds biases that may not align with business or client objectives. These constraints necessitate a prudent approach in employing GenAi, especially in applications entailing sensitive information input or leveraging the model in crucial decision-making processes.
Public Sector Applications
The applications within the public sector, particularly in public health preparedness and emergency management, are nascent but promising:
Scenario Creation: Generate realistic crisis scenarios for training and preparedness. Augment this with AI avatars, videos, and text to speech for interactive experiences.
Public Communication: Automate generation of public health messages, emergency alerts, or informational campaigns. Use AI agents to monitor the internet, compile, analyze, and interpret data (such as sentiment of current crisis),
Decision Support: Augment data analysis for informed decision-making during crises.
Resource Optimization: Enhance resource allocation strategies through data-driven insights.
Document Automation: Expedite creation of situational reports, grant proposals, and other essential documents.
Document Inquiry: Deploy chatbots to enable real-time conversational inquiries across your documents (a few or even 100s) such as emergency response plans or regulatory documents.
Skillset for Navigating Generative AI
For effective utilization of generative AI:
Grasp foundational concepts like machine learning, neural networks.
Hone data skills for quality training data preparation.
Gain hands-on experience with AI toolkits, applications, and platforms.
Leverage strategic thinking on how to integrate AI into existing workflows.
Learn to critically assess AI systems for bias, fairness, compliance, transparency, accuracy, and safety.
Are You AI Ready?
As organizations within publjc sector, public health preparedness and emergency management consider the adoption of AI, a thorough assessment of readiness is needed.
This section outlines some (but it all) critical questions to evaluate the individual and organizational capacity for responsible AI use, spanning knowledge, policy compliance, clarity of AI objectives, tool appropriateness, transparency, data privacy, bias mitigation, crisis response, and monitoring and evaluation.
Does you or your organization possess the requisite expertise in AI and related domains?
Is there a education strategy in place for enhancing AI literacy across the workforce?
Are there established policies governing AI usage within your organization?
How does your AI agenda align with existing local, state, and national legal and regulatory frameworks?
AI Objectives Clarity:
Have you clearly articulated the objectives you aim to achieve through AI
How do these objectives align with your broader organizational mission and the public good?
Is the AI tool being considered the most appropriate for achieving your stated objectives?
Have alternative technologies and solutions been thoroughly evaluated?
Does the tool provider have a history with public sector use?
Transparency of Use:
How will you ensure transparency in AI deployment and usage?
Is there a strategy for communicating AI applications to stakeholders and the populace?
Data Privacy Considerations:
How will data privacy be ensured in your AI initiatives?
Are there mechanisms to protect sensitive and personal information?
What steps will be taken to identify and mitigate biases in AI systems?
How will inclusivity and fairness be promoted in AI deployment?
Have you considered how you'll respond to challenges for AI use records?
Cost and Sustainability:
Have you considered the initial and ongoing costs of the AI tool? Is the cost fixed or scaled with use (some tools are fixed to a certain number of API calls, then priced per call. Increased use during a crisis could dramatically increase cost).
Have you envisaged potential crises that could arise from AI utilization?
Is there a robust crisis response plan to address AI-induced challenges?
Monitoring and Evaluation:
How will the impact of AI on organizational processes and public engagement be assessed?
What metrics will be employed to gauge success and areas of improvement?
A broader analytical framework and toolkit is needed to spur a nuanced deliberation on the readiness and the ethical considerations vital for responsible AI adoption.
Keep an eye out for some new tools and services to help with this, including:
a self diagnostic tool
a template based group exercise/discussion guide
a facilitated exercise/discussion service, with findings and recommendations
an e-book - Are You AI Ready?: Critical Questions Every Public Sector Leader Should Ask Before Using AI