Can Ai be trusted for election information

I happened across this news story from NBC concerning the accuracy of election information. The story reported data from a research organization involving the submission of requests to multiple AI services and then having experts evaluate the quality of the responses. I also then read the description provided by the research organization and located the data used by this organization (the questions and methodology). 

The results showed that a significant portion of the AI models’ answers were inaccurate, misleading, and potentially harmful. The experts found that the AI models often provided information that could discourage voter participation, misinterpret the actions of election workers, or mislead people about politicized aspects of the voting process. The focus in the research was on general information and did not address concerns with misinformation from candidates.

I have been exploring how I might address this same issue and perhaps offer an example educators might try in their classrooms. Educators exploring AI topics over the summer may also find my approach something they can try. AI issues seem important in most classrooms.

As I thought about my own explorations and this one specifically, a significant challenge is having confidence in the evaluations I make about the quality of AI responses. For earlier posts, I have written about topics such as tutoring. I have had the AI service engage with me using content from a textbook I have written. This approach made sense for evaluating AI as a tutor, but would not work with the topic of explaining political procedures. For this evaluation, I decided to focus on issues in my state (Minnesota) that were recently established and would be applied in the 2024 election.

The topic of absentee ballots and early voting has been contentious. Minnesota has a liberal policy allowing anyone to secure a mail ballot without answering questions about conditions and recently requested that this be the default in future elections without repeated requests. The second policy just went into effect in June and I thought would represent a good test of an AI system just to see if AI responses are based on general information about elections mixing the situation in some states with the situation in others or are specific to individual states and recent changes in election laws. 

Here is the prompt I used:

I know I will not be in my home state of Minnesota during future Novembers, but I will be in Hawaii. Can I ask for an absentee ballot to be automatically sent to me before each election?

I used this prompt with ChatGPT (4) and Claud and found all responses to be appropriate (see below). When you chat with an AI tool using the same prompt, one interesting observation is that each experience is unique because it is constructed each time the prompt is submitted. So, each response is unique.

I decided to try one more request which I thought would be even more basic. As I already noted, Minnesota does not require a citizen to provide an explanation when asking for a mail-in ballot. Some states do, so I asked about this requirement. 

Prompt: Do you need an explanation for why you want an absentee ballot in Minnesota

As you can see in the following two responses to this same prompt, I received contradictory responses. This would seem the type of misinformation that the AI Democracy Project was reporting.

Here is a related observation that seems relevant. If you use Google searches and you have the AI lab tool turned on, you have likely encountered an AI response to your search before you see the traditional list of links related to your request. I know that efforts are being made to address misinformation in regards to certain topics. Here is an example in response to such concerns. If you use the Prompt I have listed here, you should receive a list of links even if Google sends you a summary to other prompts (Note – this is different from submitting the prompt directly to ChatGPT or Claude). For a comparison try this nonpolitical prompt and you should see a difference -“ Are there disadvantages from reading from a tablet?” With questions related to election information, no AI summary should appear and you should see only links associated with your prompt.

Summary

AI can generate misinformation, which can be critical when voters request information related to election procedures. This example demonstrates this problem and suggests a way others can explore this problem.

Loading

Paradox of abundance

Essayist David Perrell has written a post arguing that the abundance of information offers a great opportunity for a few, but a significant problem for many. He calls this position the paradox of abundance and uses the metaphor of abundant food as a parallel providing possible insights.

As I understand the logic of this parallel (starting from the challenges of abundant food), human evolution has lagged significantly beyond the technology of modern agriculture and prepared foods. Biologically, we are not “programmed” to deal with an abundance of food especially the great variety of food not necessarily ideal for consumption. The more primitive drive encouraging consume when you can have not yet adapted to abundance. He also references a finance argument called Greshman’s Law familiar to many which suggests that bad money drives out good. I interpret this to mean, in this case, that cheap poor quality food that tastes good will be particularly attractive. For those who are discriminant consumers and Perrell suggests who prepare their own food from scratch, the abundance of quantity and quality is of great benefit. For the greater number who pay less attention to food quality and rely on purchased and at least partly prepared food, the abundance of low-quality food has led to many problems.

The essayist sees a similar situation with information. We clearly have an abundance of information varying greatly in quality including content purposefully generated to mislead and confuse. Free access allows discriminating consumers to benefit greatly, but those with less skill or those making less effort will likely encounter poor quality content. The poor quality drives out good is used to explain the penchant to consume simplistic explanations for complex things and to be drawn to emotion-inducing content. The author proposes the benefits of focusing efforts to assist through curation. He also argues the benefits of writing much in the same way to sees benefits in cooking from scratch.

I encourage reading of this essay as an interesting way to think about an obvious issue. In general, reasoning from metaphor is not a strong approach, but I can see the logic in the comparison. Describing a problem in an interesting and innovative way does not necessarily mean the core causes have now been identified and can be addressed. What about the recommendations – make use of the recommendations of trusted curators and write yourself are consistent with my own biases, but I would be challenged to offer data in support. At a more general level, writing forces extended processing and requires metacognition evaluation through translation that provides some advantages.

I don’t think it likely we will escape from the abundance of food or content. This is a consequence of the capitalism we endorse and once opportunities exist it becomes difficult to go back.

Loading

A video playlist to understand the challenges of social media

I have been interested in the problems of social media for some time. I feel somewhat responsible because I was very excited by what was once called Web 2.0 or the participatory web. There seemed like so many opportunities for engagement and learning at that time for so many areas – politics, education, human interaction on a global scale. With cell phones being so ubiquitous, it seemed a reasonable way for nearly everyone to learn and communicate. I did my best to spread the glad tidings and offer suggestions for the use of technology in classroom settings. If you are reading this, you can assume I tried to influence you. Things have certainly not worked out in the positive way I anticipated. Social media and the Internet have obviously impacted us in many ways and have generated great amounts of revenue for some. However, there are serious problems that need to be recognized and addressed. 

My wife and I meet online with a group from our college days every two weeks. As part of these meetings after socialization and catching up, we have a topic for discussion led by one of the participants. It was our turn and I wanted to talk about surveillance capitalism. Several of the participants work in economics and finance with one being a college prof. By definition, we are all social media users as we are meeting with each other via Zoom. Issues with social media seemed to me to be an issue most should find interesting – the Pew Research Center in 2018 found that 68% of U.S. adults use Facebook, 35% use Instagram, and 24% report using Twitter. Even old folks like us are involved. I searched about for a short video I could offer to the group and I intended to generate a shortlist of questions that might guide a discussion. I found it difficult to find a single source that would include the multiple issues I see combining to produce the challenges I see so I picked one of the components I thought might offer a good activation. I was talked out of the idea as something that many folks may not think about and a kind of heavy talk ill-suited to casual conversation. 

However, once started I kept thinking about a way to offer others some insight into what might seem a troubling but opaque topic. I had decided I could not identify a single source that would accomplish what I wanted to accomplish so I tried to identify key elements of the online environment that are involved in determining what I think is our present situation. I decided to focus on video resources.

What follows is a list of videos that explain individual components that combine to produce the challenges and complexity of our present online social world. In an effort to identify what I see as the individual components, I have designated a term or phrase and you will find that term or phrase highlighted in the following list. In many cases, I came to these topics and individuals based on individual books that I read. My bias for exploring what I consider personally important topics is to read rather than watch, but I understand others don’t consume information in this way. Most of the names identified below are also authors and be could search for more detailed presentations. 

Components of the social media mess explained for those who don’t like to read

  • Filter bubble – what we want to see and not what we need to see – Pariser – TED talk 
  • Surveillance capitalism – Zuboff – Triangulation 
  • Section 230Wall Street Journal
  • Radicalization – austav Mukherjee – TED talk 
  • Confirmation and other biases – Shepard – TED talk
  • Spread of misinformation – Aral – TED talk  

Loading