AI is changing the game - Are you on board?
The Exciting Future of AI and PE: Let's Dive in!
“Canada's privacy watchdog concerned over potential privacy breaches in AI-powered healthcare services.”
〰️
“Canada's privacy watchdog concerned over potential privacy breaches in AI-powered healthcare services.” 〰️
“United Nations partners with AI startup to launch an interactive platform for public engagement on climate change.”
〰️
“United Nations partners with AI startup to launch an interactive platform for public engagement on climate change.” 〰️
UK Government plans to use AI to enhance public engagement in policy-making processes.”
〰️
UK Government plans to use AI to enhance public engagement in policy-making processes.” 〰️
Reading these recent news headlines, I can’t help but ask …
Should I be looking at integrating AI in the work I do?
Should I know more? Where do I start?
What considerations are there for the use of AI in public engagement?
In this month’s Daring Dialogue we are tackling a topic that is on everyone’s mind (or at least their newsfeed) these days: Artificial Intelligence (AI)
Testing the waters
... a brief true story about our first foray into the AI world.
We recently completed a massive engagement process with thousands and thousands of lines of data. It was generated by hundreds of passionate, caring and deeply impacted community members. We were navigating tough issues of historical conflict, safety fears and discrimination. So first, we did it the “old fashioned way” – with a number of data analyzers (to check assumptions and accuracy) looking for common themes, tensions and nuances of the stories, experiences and insights provided by community members. We then painstaking reviewed the codes and themes to prepare a summary report that could be confidently shared with community members and decision-makers – one we felt confident both reflected and honoured what we heard and what people wanted decision-makers to know.
With AI being such a hot topic right now, we couldn’t help ourselves …. we wondered: “Could we have saved ourselves hundreds of hours by leaning into AI-powered analysis only to end up with the same result????” So we grabbed a big chunk of the data, ran it through ChatGPT – just to satisfy our curiosity. Now just to be clear, we did this only once, without the knowledge of how to “train” AI, and it was only with one section of data …. so more testing would be useful for proper comparison.
The end result …. Chat GPT did not bad.
It hit the mark on most of the high level, easy to find themes.
It definitely stood out in the simple and plain language it used to produce the summary
It was quick, soooooo quick
BUT it missed the really important, harder to find stuff. The nuanced themes connected to fairness, equity and compassion.
So I’ll be honest – I’m intrigued.
Let’s Learn Together
With so much information out there to digest on this topic, I want to approach the conversation a little differently and invite all of you to learn alongside me.
Clearly, I am not an AI expert. But I am curious about how AI might enhance the work I do, hopefully adding to the connectivity and impact we all seek in our engagement work. Keeping in mind my relative newness to this subject area and keen desire to discover how it might apply to the world of public engagement, I put my focus on:
My top 4 questions about AI right now
Is AI’s only role within public engagement as a tool for managing submissions and communications?
Can AI remove bias from public engagement data analysis?
How does AI account for power imbalances in public engagement?
How can AI increase accessibility in public engagement?
The top 5 AI tools that might be worth exploring
*In the spirit of full transparency, and as an exploration of the subject in question, we didn’t interview a human AI expert for our top questions or tools. We queried several AI programs themselves and read a whole lot of articles! The questions, however, are entirely ours. We’re excited to share what we found!
1. Is AI’s only role within public engagement as a tool for managing submissions and communications?
AI as a tool for Qualitative Analysis & Beyond
Well, I’m not sure yet :) We are still exploring ... and in the research we’ve done so far, we have seen that AI can certainly analyze large amounts of data and identify patterns, it’s true. Here are some of the things that AI is already doing that are exciting:
AI can provide insight into the dynamics of emerging disinformation and help identify potential rapid disinformation attacks. [example here]
Natural Language Processing (NLP) can be used to automatically analyze and extract insights from unstructured data. Automated coding can be used to automatically code qualitative data by identifying and tagging specific themes or topics. This can help speed up the analysis process and reduce the risk of human error. [see our AI tools section at the end of this email for examples]
Sentiment analysis: AI-powered sentiment analysis can be used to automatically classify opinions expressed in qualitative data as positive, negative, or neutral. This can help identify trends and patterns in public sentiment and inform decision-making processes. [more on that here]
Topic Modeling: AI can be used to automatically identify and extract topics from large amounts of qualitative data, which can help categorize and organize the data in a meaningful way. [more on that here]
2. Can AI remove bias from public engagement data analysis?
AI as a Safeguard Against Bias
Could you imagine??!! PE Practitioners are constantly assessing and reflecting on our own bias – doing our best to dig up those deep seated unconscious biases in an attempt to serve and support the tough conversations we navigate. We know there is bias in the data, we also know that as data analyzers, we bring our own world view. So, while we do our very best to recognize and hold those judgements and assumptions at bay I imagine that they do impact our analysis.
So from what we’ve researched so far, AI has the potential to reduce bias in public engagement data analysis, but it cannot completely eliminate it. AI algorithms are only as unbiased as the data they are trained on. If the training data contains biased or incomplete information, the AI model will also produce biased results.
That being said, there are measures that can be taken to reduce bias in AI-based public engagement data analysis. These include:
Ensuring that the training data is diverse and representative of the population being studied.
Regularly reviewing and auditing AI algorithms to detect and mitigate bias.
Incorporating ethical considerations into the development and deployment of AI systems.
Conducting user testing and incorporating feedback from diverse stakeholders.
3. How does AI account for power imbalances in public engagement?
AI as a Champion of Underrepresented Peoples
AI may not inherently account for power imbalances in public engagement, but it can be designed and used in the following ways that mitigate or address such imbalances.
Ensuring diversity and representation: AI can be used to analyze participation patterns and identify groups that are underrepresented or marginalized. This can help organizations take proactive steps to ensure that diverse perspectives are included and accounted for in public engagement processes.
Incorporating ethical considerations: AI developers and users can incorporate ethical considerations, such as fairness, accountability, and transparency, into the design and deployment of AI systems.
Providing alternative engagement channels: AI-powered tools like chatbots can be used to provide alternative engagement channels that may help increase participation and reduce barriers to engagement.
4. How can AI increase accessibility in public engagement?
AI as an Instrument for Greater Accessibility in Engagement
Natural language processing (NLP) can be used to automatically transcribe speech into text, or to translate written text into different languages, making public engagement materials more accessible to a wider audience.
AI-powered chatbots and virtual assistants can provide instant responses to questions and inquiries, 24/7, which can improve accessibility for individuals who may not have access to human representatives or who prefer self-service options.
Predictive analytics: AI can be used to analyze large amounts of data to identify patterns and predict user behavior, which can improve accessibility by helping organizations anticipate and respond to the needs of their audience.
Let's Talk AI Tools
The following is a list of tools that we discovered during our exploration of the topic. Although we can't personally attest to their applicability and usefulness in every area of public engagement, we can use them as a starting point.
Outwrite: a paraphrasing tool that uses AI to rephrase sentences and paragraphs to improve writing clarity and readability. Best of all you can view & adjust the grade level! Website:
NVivo: A qualitative data analysis tool that allows users to organize, analyze, and visualize unstructured data. It offers features for coding, content analysis, and text search.
Atlas.ti: Another qualitative data analysis software that allows users to analyze and visualize text, audio, video, and image data. It offers features for coding, content analysis, and network analysis.
Rose AI: Rose in an AI tool for finding and visualizing data. Just type in what you’re looking for and it will generate a gorgeous infographic for you.
Perplexity: Perplexity is a lot like ChatGPT but with more refined answers and cited sources.
BONUS!
Two articles + 1 video we found interesting
Artificial Intelligence: Real Public Engagement
This is a super interesting article about the work RSA is doing to explore if we can ethically engage citizens and AI for automated decision-making. Fascinating!!!
Here is a mini guide to getting started using ChatGPT in an informed way!
I hope you found value in this month's Daring Dialogue topic!
Let me know if you've tried any of these tools! Have you asked yourself different questions? What answers can you share?
Your partner in learning - Kim