Dear Friends,
This last weekend I spent my time installing a second bee hive. I became a bee-keeper three years ago, and since last fall I’ve been dreaming of expanding my backyard endeavors. Last Saturday was pick-up day, a date determined by completion of almond blossom pollination in California, and thousands of bees needing to find new homes across the country. To say it must have been a shock for them is an understatement. I collected a package of 10,000 (3lbs) bees and a new queen, which had travelled 1,000+ miles, then set about installing them amidst freezing temperatures and 5 inches of snow! No, not ideal for starting a new colony (thank you springtime in Boulder, CO!). Thankfully, even with these extreme circumstances they appear to have not only survived, but are thriving. A message of resilience and perseverance to us all when we come together as a community!
It’s a memo which we can’t remind ourselves of often enough, for it’s been yet another challenging week for civil society. In one short week a former US president became the first holding such an office to stand trial in a criminal court, the Supreme Court heard arguments in a case (Grants Pass v. Johnson) that could allow cities to criminalize the homeless for sleeping on the streets, the House finally approved a bi-partisan bill of $95 billion aid bill for Ukraine, Israel and Taiwan supporting aid efforts for wars on multiple continents. Oh and at the same time managed to slip in legislation that could trigger a nationwide ban of TikTok if its Chinese owner does not sell the video app. We also marked the 25th anniversary of the Columbine school shooting here in Colorado; with the mass shootings still making daily headlines, and gun safety an ever present civil debate.
These topics have made my own head spin, and they were only some of the critical topics of the last seven days. Any wonder we are having such a hard time keeping up, never mind trying to engage on every topic which demands our attention.
As I move forward with “Common Ground” I am realizing that part of what I want to do is to decouple civil discourse away from the media and the headline grabbing stories. Each night we can lament the latest events of the day, be outraged and despondent, but rarely are we motivated to learn more or take action. We need to slow down to move faster.
How would things change if we had an opportunity to deeply discuss and understand the nuance of issues which quite rightly are too big to consume in bit size pieces or headlines.
Today, civic engagement is inches deep and miles wide, and that needs to change.
We’ve long felt frustrated and shackled, and unclear how to participate in the democratic process beyond elections. The big question is how? And that’s where I’m focusing my attention.
Next week I’ll be sharing more about what I’ve been learning and how “Common Ground” could evolve. I’m excited and also realizing the scale in which we need a revival of civic engagement in order to create a positive future for everyone.
The good news is that as a society we are more aware of the fragility of democracy than at any time before in human history. That’s a massive opportunity to gather and organize.
Hold that thought!
Follow Up to Last Week’s Poll
Before we get into this week’s topic, I wanted to check back in on the question we addressed last week on "Where should we have important conversations online?" Thank you to the few hand full of readers who completed the Substack poll. The results while only a small study, reflect what we are all seeing - there is indeed no one good place to discuss the topics of civic society. LinkedIn and ‘Other’ came in at 38% and Reddit at 25%. No one chose Facebook or “X”. Sadly due to the nature of Substack, folks couldn’t complete which ‘other’ they would have suggested. Next time, I’ll be sure to use a different poll/survey tool. Rookie mistake. If you did respond with other… please hit reply and let me know what you would suggest.
From civil society technology challenges to efforts which might give us more cause to hope.
The Role of AI in Civil Discourse
We spend a bunch of time at home talking about AI. Pascal’s (my better half) work on disruption has brought him ever deeper into this new realm with business audiences wanting to know how they can best prepare for, and leverage this new frontier. Probably much like yourselves, we go back and forth between AI’s vast potential, and a Black Mirror style dystopian future. We usually land in the glass half full arena (our typical MO) with a plethora of rapid developments and increasingly exciting use cases.
Whilst AI is front and centre in the mind of corporate America, I have also been heartened to see civil society embrace it’s potential capabilities for good. So even though today social media appears to be hindering our ability to have important conversations online, it evident some technology could prove in aiding balanced and modulated discussion, and information sharing.
AI-Driven Interventions Improve Discussion Quality
In the last few years several studies have been conducted using AI for equitable civil discourse, working to ensure transparent, fair, and healthy dialogues with encouraging results.
In one study from 2023 entitled “Leveraging AI for democratic discourse: Chat interventions can improve online political conversations at scale” by Duke University, and Brigham Young University, researchers used OpenAI’s GPT-3, a cutting edge generative AI language model at the time (GPT-5 is reported to be released this Summer), to provide participants with real-time suggestions whilst in conversation. The goal was to enhance each person's sense of being heard and understood by their conversation partner. Remarkably, these generative AI-driven interventions led to notable improvements in the overall quality of the discussions, as reported by the participants themselves.
As written about in BYU’s Daily Universe, part of the participant study included using GPT-3 to make their online political dialogue more civil around the divisive issue of gun control. “Participants would receive suggestions on what words they should rephrase so other participants felt more understood. For example, rather than someone who supports stricter gun control regulation saying “guns are a stain on democracy,” the GPT-3 software would suggest “I understand that you value guns.””
The AI prompts such as this encouraged a more nuanced and respectful exchange between individuals with differing viewpoints, where both parties felt equally valued and respected. The tone of the conversations also became more positive and constructive, without the AI altering the substance of the dialogue or unduly influencing their personal beliefs.
So while concerns persist about the potential risks associated with AI in exacerbating social divisions such as this latest research from Anthropic on it’s increasingly “persuasiveness” language models. Studies like that from BYU and Duke start to underscore the potential for it to play a much bigger supportive role in fostering more meaningful and productive political conversations.
AI's Role in Gauging Voter Sentiment and Predicting Results
Large scale technology systems have already been widely used for some time to gauge voter sentiment and predict results through election polling. By analyzing diverse data sources, the hope is that AI systems can also provide a more accurate representation of public sentiment, going beyond the loudest voices in the room and removing bias. This is essential for ensuring that the needs and opinions of all people are considered, while still respecting the autonomy and agency of participants involved.
In 2020, several AI firms correctly predicted results for the US presidential contest, including in 11 battleground states and by what margin. That being said, there were also a bunch of 2020 AI generated poll predictions with plenty of errors. The technology back then was still in its infancy. Four years on with a new US election looming, we are certainly going to see some fascinating forecasts. Of course, the devil will be in the details to see how far we’ve come to accurately predict voting behavior.
Staying on the election topic, as concerns increase over the rise of deep fakes and disinformation, particularly this year as half the world will cast a vote, so too could AI work to help combat these threats. As FBI Director Christopher Wray has said “In some ways, A.I. is quite good at detecting A.I.”
AI’s applications hold the potential for both good and harm. We’ve been here before, we must ensure good wins out.
The Need for Careful Development and Oversight of AI Systems
One of the fundamental principles of AI is that they can’t operate in right or wrong, rather in more or less likely outcome. This means they have a vast potential for aiding in civil discourse and sentiment analysis. The very structure of these large language models (LLM) inherently lend themselves to promoting impartial dialogue and analysis. That’s why we need to be ever more careful about how they are trained. Realizing the potential of AI in the civic realm is far from conclusive. It will require a vast commitment and overseeing of developing AI systems that prioritize fairness, and healthy, democratic values, whilst removing bias.
Though caution is certainly warranted, I believe we can be optimistic about the role AI can play in enhancing our capacity to connect, understand and learn from one another. It has tremendous promise in its ability to tackle some of our most intractable challenges. I’ll be sure to post further as more studies and use cases come to light.
“Meet in the Middle”
This week a simple challenge. Have you used/interacted with an AI yet?
As responsible members of society, it is crucial that we stay informed about emerging technologies. Given the rapid advancement and increasing prevalence of AI, I believe everyone should have a basic understanding of its potential and limitations.
Claude.ai, ChatGPT, Copilot - there are so many to choose from. The important piece is to spend the $20 (average monthly cost) to use the latest version of the technology. It’s the difference between having a conversation with a young teenager and someone much more experienced. No offense to teens who are honestly smarter than a bunch of adults these days!
Why not try asking the AI you choose what it thinks about a civil society topic such as defunding the police, banned books, border control. I think you’ll be surprised by the answers you receive.
As always, if you feel comfortable, I'd love to be in touch and hear your observations, reflections and questions. Feel free to email me at jane@finette.com or share your thoughts in the Substack comments.
Remember - there's always more that unites us than divides us.
Let's find common ground together,
Jane