AI – Page 2 – Learning Aloud

Upload Obsidian Folders to NotebookLM

Many of us have now had the opportunity to spend considerable time with AI tools and have sorted out for ourselves what types of applications are most productive. For my personal use, I found the greatest value in the use of AI to analyze the notes and highlights I have accumulated over the years to search for insights I might use in my writing projects. I keep most of my notes in Obsidian and one of the more popular posts I have generated focusing on AI and personal notes explained my approach to using the Obsidian AI extension Smart Connections.

As I continue exploring, I have found another approach that others may find useful. This approach uses the Google AI tool NotebookLM. This tool is presently provided at no cost and is easy to use. There is a frustrating mismatch between how most of us use Obsidian and the structure of NotebookLM. Obsidian users typically create many short notes and then create links and tags to allow these notes to be stored and used in the most flexible manner possible. NotebookLM allows the uploading of considerable content into an individual notebook to be explored with AI chats, but limits the number of files that can be included. This combination – many small files in Obsidian and a limited number of files in NotebookLM notebooks – creates a challenge.

Here is my present solution. There is an extension for Obsidian (Better Export PDF) that allows the content from multiple files stored within an Obsidian folder to be exported as a single PDF. You can probably see where this is going. Export the many files in several folders as PDFs and then add these PDFs to NotebookLM.

The Better Export PDF in Obsidian

The following image was captured as I used Obsidian. You can see the multiple folders I have created. After One Better Export has been added to Obsidian, you first right-click (control click) on a folder to bring up various options. The one you want is Export folder to PDF.

Activating this option may leave you confused as the process takes some time to accomplish. You know when things are working properly when you see the names of individual files appear towards the top of the display (red box). When the process of concatenating the files has been completed the complete PDF file will appear in what initially appears as white space. Now you export this large file.

You complete this process additional Obisian folders until you have exported the large PDFs you want to explore in NoteBookLM.

NotebookLM AI Chat

The following image shows a notebook I have created based on three PDFs (Cognitive, Generative, Notetaking). A couple of hundred individual notes are included in these three large files. At the bottom of the image, you should find a prompt I have entered.

For AI tool users, everything should now be more familiar. NotebookLM has responded to my original prompt and I can accept some of the recommendations for related topics I might explore (not shown). NotebookLM offers a way to review the content the AI used to generate the prompt reply (numbers in the red box). Selecting one of these numbers takes you to the section in your input and allows you to explore that context.

Summary

So, this combination of Obsidian and NoteBookLM seems to work well. The key is the Obsidian extension Better Export PDF which allows many Obsidian notes to be combined to address the limited number of files NoteBookLM will accept as input. Give it a try.

ChatLLM from Abacus AI

I have this issue with AI services. I have identified several services that provide me useful services (image generation for posts, edits for what I write, identification of journal articles relevant to my writing, and chats with my large collection of notes and highlights) The issue is no individual tool does a great job of all of these things and I don’t use any of the several individual tools enough to justify the $20 a month that seems to be the going rate for any given tool. I could keep telling myself that this retirement writing gig should be treated as a hobby that I can afford, but there is a principle involved here. I think of many of my posts as exploring tools for teachers and $75-100 a month in their situation is likely a bit much.

I have found a general-purpose AI service, the ChatLLM tool from Abacus AI, that I have for me what is unlimited use for $10 a month. For a month or so now, I have been duplicating my use of ChatLLM and ChatGPT and can find very little difference in what the tools generate. Abacus.AI has an introductory offer that is a little confusing. The first month is free, but you must purchase one month and then you can cancel after the first two months.

The interface is very similar to what you experience with other AI services. There is a prompt field and responses appear above. ChatLLM allows the individual series of chat interactions with a given purpose allowing you to return if you have a reason to pick up again on an earlier exploration.

ChatLLM appears to function as an API allowing you to select from different AI services. This option seems useful for those of us who like to explore the possible differences among the AI services that are available.

AbacusAI offers many features I have not bothered to explain here in detail. I have learned what I thought necessary to meet my personal needs and it should be very easy for anyone with any experience with other platforms to get started. Here is a more complete review by another writer.

How many AI tools?

It’s not that I don’t find AI to be useful. I generate a half dozen images a month to embellish my writing. I search for journal articles I then read to examine an educational issue I want to write about. I examine what I have written to identify errors in grammar or syntax or even identify my use of passive voice which I still can’t figure out. My issue is the monthly subscription fees for the multiple tools that best suit these and other uses. It is simply difficult to justify the $20 a month fee which seems to be the going rate for each of the services and the level of use I make of each service.

I regard my use of AI as both a benefit to personal productivity, but also a subject matter I explore and write about. Writing about the intersection of technology and education is a retirement hobby and I don’t need to do what I do on a budget. However, I don’t think this is true for everyone and I can set as a personal goal an exploration of the financial issues others might need to consider. AI tools differ in how flexible they are. Often, the less flexible tools are optimized to accomplish a specific set of tasks and this narrow range increases ease of use. Users can find themselves evaluating the cost-effectiveness of options based on ease of use versus total cost.

Grammarly as an example

As someone who spends a great deal of time writing, it is worth my time to consider how AI tools can be used to improve the productivity of the time I spend writing and the quality of what I write. Grammarly is a tool suited to such goals. I have relied on both the free and pro versions of Grammarly for a quick way to edit my writing. This is a great product I find helpful. Again, the issue is the Pro version adds another $12 to the total of my monthly subscriptions bill. The free version is great for basic errors, but the pro version is necessary for complete sentence rewrites. It is helpful to compare what I said with the suggested rewrite to decide whether I want to accept the alternate version.

What I have been exploring lately is the use of more general-purpose AI tools (e.g., ChatGPT, Abacas.ai) as an alternative. I understood that many others were using such tools in this matter, but because I had already been using Grammarly I had not invested the time to explore how more flexible tools might be used to achieve similar goals. Here is the approach I have settled on. I use a multi-level prompt to specify what I want to learn from an evaluation of something I have written. I keep this prompt in a note app so that I can copy and paste the prompt at the beginning of the content I want to evaluate. The prompt appears below. The prompt is written to be flexible so that I can adapt it to different circumstances by adding or removing specific guidelines. For example, I realize I often write in passive voice and some tasks should not include text expressed in this way. The guideline “write in active voice” can be included when I think this approach is important. If writing for a specific age group might be a goal, a guideline related to this goal can be inserted.

To demonstrate what this prompt accomplishes, I will take what I written here to this point and apply the prompt.

Evaluate the following content using the following guidelines:

Correct spelling errors
Correct grammatical errors
Correct punctuation errors
Write in active voice
Suggest corrections
Only list the lines containing errors and the proposed corrections
Add an explanation for the need for a correction

[Insert content to be evaluated here]

Response from Abacas.ai.

1. **Original:** I examine what I have written to identify errors in grammar or syntax or even identify my use of passive voice which I still can’t figure out.

   **Correction:** I examine what I have written to identify errors in grammar or syntax and even identify my use of passive voice, which I still can’t figure out.

   **Explanation:** Added a comma before “which” to correctly punctuate the non-restrictive clause.

2. **Original:** It is simply difficult to justify the $20 a month fee which seems to be the going rate for each of the services and the level of use I make of each service.

   **Correction:** It is simply difficult to justify the $20-a-month fee, which seems to be the going rate for each of the services and the level of use I make of each service.

   **Explanation:** Added a hyphen in “$20-a-month” to correctly form a compound adjective and a comma before “which” to punctuate the non-restrictive clause.

3. **Original:** I regard my use of AI as both a benefit to personal productivity, but also a subject matter I explore and write about.

   **Correction:** I regard my use of AI as both a benefit to personal productivity and a subject matter I explore and write about.

   **Explanation:** Replaced “but also” with “and” to correct the parallel structure.

4. **Original:** Writing about the intersection of technology and education is a retirement hobby and I don’t need to do what I do on a budget.

   **Correction:** Writing about the intersection of technology and education is a retirement hobby, and I don’t need to do what I do on a budget.

   **Explanation:** Added a comma before “and” to separate two independent clauses.

5. **Original:** Grammarly as an example

   **Correction:** Grammarly is an example.

   **Explanation:** Added “is” to form a complete sentence.

6. **Original:** Again, the issue is the Pro version adds another $12 to the total of my monthly subscriptions bill.

   **Correction:** Again, the issue is that the Pro version adds another $12 to the total of my monthly subscription bill.

   **Explanation:** Added “that” for clarity and changed “subscriptions” to “subscription” for correct singular form.

7. **Original:** What I have been exploring lately is the use of more general-purpose AI tools (e.g., ChatGPT, Abacas.ai) as an alternative.

   **Correction:** What I have been exploring lately is the use of more general-purpose AI tools (e.g., ChatGPT, Abacus.ai) as an alternative.

   **Explanation:** Corrected the spelling of “Abacas.ai” to “Abacus.ai.”

8. **Original:** I understood that many others were using such tools in this matter, but because I had already been using Grammarly I had not invested the time to explore how more flexible tools might be used to achieve similar goals.

   **Correction:** I understood that many others were using such tools in this manner, but because I had already been using Grammarly, I had not invested the time to explore how more flexible tools might be used to achieve similar goals.

   **Explanation:** Changed “matter” to “manner” for correct word usage and added a comma after “Grammarly” to separate clauses.

The feedback I have received would normally have resulted in my making several adjustments. I have not done so for you to consider both the original and the suggestions. The feedback makes specific suggestions and I either copy and paste the suggestions to the original source material or simply retype what needs to be fixed. This is clumsier than just clicking on the recommendation offered by Grammarly and it often requires that I use the search function in a longer text to locate the offending sentence. When writing on my desktop machine I simply open a second window and move between one window with the AI feedback and the original document to make adjustments.

Potential conflicting benefits of your note-taking tool and approach

As I have explored and used several digital note-taking tools and examined the arguments that have been made regarding how such tools result in productivity benefits, I have identified a potential conflict in what produces more positive outcomes. The recognition of this conflict allows more purposeful execution on the part of the tool user and may better align activities with goals.

One way to identify note-taking goals is to use a long-standing approach differentiating generative and external storage benefits. This distinction was proposed long before PKM and was applied in the analysis of notes taken in classroom settings. The generative benefit proposes that the process of taking notes or sometimes of taking notes in a particular way engages our cognitive (mental) processes in ways that improve retention and understanding. External storage implies that our memory becomes less effective over time and having access to an external record (the notes) benefits our productivity. In practice (e.g., a student in a classroom) both benefits may apply, but one benefit depends on the other activity. Taking notes may not be beneficial, but to review notes one must have something to review. This is not always true as notes in one form or another can be provided or perhaps generated (for example AI identification of key ideas), but taking your own notes is by far the most common experience. In a PKM way of thinking, these two processes may function in different ways, but the classroom example should be familiar as a way to identify the theoretical benefits of note-taking.

I have written about the generative function of note-taking at length, but it is important to point out some unique specifics that apply to some digital note-taking tools. A source such as Ahrens’ Taking Smart Notes might provide the right mindset. I think of generative activities as external actions intended to produce a beneficial mental (cognitive) outcome. The idea is that external activities can encourage or change the likelihood of beneficial thinking behaviors. One way of operationalizing this perspective is to consider some of the specific activities Ahrens identified as external work resulting in such cognitive benefits. What are some of these activities? Isolating specific ideas and summarizing each as a note. Assigning tags that characterize a note. Making the effort to link notes. Periodically reviewing notes to generate retrieval practice, to reword existing notes, and to add new associations (links).

Retrieval is easier to explain. Note-taking apps with highly effective search capabilities make it easy to search and surface stored information when it might be useful. Links and tags may also be useful in this role, but search alone will often be sufficient.

What about the potential conflict?

The conflict I see proposes that some tools or approaches rely more heavily on search arguing in a way that generative processes are unnecessary.

I starting thinking about this assumption when contrasting the two note-taking systems I rely on – Mem.ai and Obsidian. While Mem.AI and Obsidian could be used in exactly the same way, Mem.ai developers argued that the built-in AI capabilities could eliminate the need to designate connections (with tags and links) because the AI capabilities would identify these connections for you. Thus when retrieving information via search, a user could use AI to also consider the notes with overlapping foci. If a user relied on this capability it would eliminate the work required to generate the connections manually created in Obsidian, but this approach would then also avoid the generative benefits of this work.

AI capabilities fascinate me so I found a way to add a decent AI capability to Obsidian. Smart Connections is an Obsidian plugin that finds connections among notes and allows a user to chat with their notes. So, I found a way to mimic Mem.ai functionality with Obsidian.

I find I have found a way to alter my more general PKM approach because of these capabilities. Rather than taking individual notes while reading, I can annotate and highlight pdfs, books, and videos and export the entire collection for each source and then bring this content into both Mem.ai and Obsidian as a very large note. Far easier than taking individual notes, but at what generative cost?

Smart Connections has added a new feature that even facilitates the use of the large note approach. Connections finds connections based on AI embeddings. An embedding is the mathematical representation of content (I would describe as weights based on what I remember of statistics). The more two notes embeddings’ weights are similar the more the notes consider similar ideas. Smart Connections used embeddings to propose related notes. Originally embeddings were generated at the note level and now at the “block” level. What this means (block level) is that Smart Connections can find the segments of a long document that have a similar focus as a selected note.

Why is this helpful? When I read long documents (pdfs of journal articles or books in Kindle), I can export a long document containing my highlights and notes generated from these documents. With Smart Connections I can then just import this exported material into Obsidian and use Smart Connections to connect a specific note to blocks of all such documents. I can skip breaking up the long document into individual notes and assigning tags and creating links.

Why is this a disadvantage? Taking advantage of this capability can be a powerful disincentive to engaging in the generative activities involved in creating and connecting individual notes the basic version of Obsidian requires.

Summary

As note-taking tools mature and add AI capabilities, it is important for users to consider how the way they use such tools can impact their learning and understanding. The tools themselves are quite flexible but can be used in ways that avoid generative tasks that impact learning and understanding. If the focus is on the retrieval of content for writing and other tasks, the generative activities may be less important. However, if you start using a tool such as Obsidian because a book such as Smart Notes influenced you, you might want to think about what might be happening if you rely on the type of AI capabilities I have described here.

References
Ahrens, S. (2022). How to take smart notes: One simple technique to boost writing, learning and thinking. Sönke Ahrens.

Show Your Work – Improve Argumentation

Thinking is not visible to self and others and this reality limits both personal analysis and assistance from others. I have always associated the request to show your work associated with learning math so subprocesses of mathematical solutions can be examined, but the advantage can be applied when possible to other processes. I have a personal interest in the ways in which technology can be used to externalize thinking processes and the ways in which technology offers unique opportunities when compared with other methods of externalization such as paper and pen.

Ideas from different areas of interest sometimes come together in unexpected ways. This has been a recent experience for me with a long-term interest in argumentation and digital tools applied to learning. Argumentation may not spark an immediate understanding for educators. It sometimes help if I connect it with the activity of debate, but it relates to many other topics such as critical thinking and the processes of science as well. It relates directly to issues such as the distribution of misinformation online and what might be done to protect us all from this type of influence.

For a time, I was fascinated by the research of Deanna Kuhn and wrote several posts about her findings and educational applications. Kuhn studied what I would describe as the development of argumentation skills and what educational interventions might be applied to change the limitations she observed. It is easy to see many of the limitations of online social behavior in the immaturity of middle school students engaged in a structured argument (debate). Immature interactions involving a topic with multiple sides might be described as egocentric. Even though there is an interaction with a common topic, participants mostly state the positions they take with and frequently without supporting evidence. As they go back and forth, the seldom identify the positions taken by an “opponent” or offer evidence to weaken such positions. Too often, personal attacks follow in the “adult” online version, and little actual examination of supposed issues of interest is involved.

Consideration of the process of clearly stating positions and evidence for and against maps easily to what we mean by critical thinking and the processes of science. In the political sphere what Kuhn and similar researchers investigate relates directly to whether or not policy matters are the focus of differences of opinion.

Externalization and learning to argue effectively

Kuhn proposed that to improve (develop) critical thinking skills learners would benefit from experiences encouraging reflection. An approach that proved productive was based in multiple studies on two techniques for encouraging reflection. Across multiple age groups (middle school, high school, college students) she had pairs of participants argue using online chat. A pair had to agree on a given “move” or statement before submission (externalizing rationales for consideration) and submitting statements in chat both allowed an opportunity to focus on the message with interference from the face-to-face issues that are present in formal debate and to create a record that could be critiqued. In some studies, the participants were asked to complete forms asking for a statement of the positions taken by opponents and evidence offered in support of these positions. The effectiveness of the treatments was examined following training without such scaffolds.

AI arguments result in an external record

I and others have been exploring the experience of arguing with an AI opponent. One insight I had while exploring this activity was that it resulted in an external product that could be examined much in the way Kuhn’s chat transcripts could be examined. Classroom applications seem straightforward. For example, the educator could provide the same prompt to all of the students in the class and ask the students to submit the resulting transcript after an allotted amount of time. Students could be asked to comment on their experiences and selected “arguments” could be displayed for consideration of the group. A more direct approach would use Kuhn’s pairs approach asking that the pairs decide on a chat entry before it was submitted. The interesting thing about AI large language models is that the experience across submissions of the same prompt are different for each individual or for the same individual submitting the prompt a second time.

I have described what an AI argument (debate) looks like and provided an example of a prompt that would initiate the argument and offer evaluation in a previous post. I have included the example I used in that post below. In this example, I am debating the AI service regarding the effectiveness of reading from paper or screen as I thought readers are likely familiar with this controversy.

…

Summary

Critical thinking, the process of science, and effective discussion of controversial topics depends on the skills of argumentation. Without development, the skills of argumentation are self-focused lacking the careful identification and evaluation of opposing ideas. These limitations can be addressed through instructional strategies encouraging reflection and the physical transcript resulting from an argument with an AI-based opponent provides the opportunity for reflection.

References:

Iordanou, K. (2013). Developing Face-to-Face Argumentation Skills: Does Arguing on the Computer Help. Journal of Cognition & Development, 14(2), 292–320.

Kuhn, D., Goh, W., Iordanou, K., & Shaenfield, D. (2008). Arguing on the Computer: A Microgenetic Study of Developing Argument Skills in a Computer-Supported Environment. Child Development, 79(5), 1310-1328

Mayweg-Paus, E., Macagno, F., & Kuhn, D. (2016). Developing Argumentation Strategies in Electronic Dialogs: Is Modeling Effective. Discourse Processes, 53(4), 280–297. https://doi.org/10.1080/0163853X.2015.1040323

Is AI overhyped? Maybe Apple has the right idea.

In my world, talk of AI is everywhere. I doubt most have a different opinion because nearly any news program has a story every other day or so commenting on AI capabilities, dangers, and the wealth and power being accumulated by developers. We all have experienced the history, beginning with ChatGPT in Nov. 2022, of the large language models.

I tried to find some specifics about the popularity of AI and this is challenging. There were quickly multiple companies involved and you can use free versions of AI programs with just a browser making an accurate of “users” difficult. We do know that a million users signed up for ChatGPT 3 within three months.

So, where are we at a year and a half later? Again, you and I may use an AI large language model on a daily or at least weekly basis, but how much use is actually going on “out there”?

Studies have started to appear attempting to determine the frequency of frequent use. Frequent use can be very different from “yeah, I tried that” use. My interpretation is that folks in many countries have heard of AI and quite a few have given at least one service a try, but most now appear puzzled by what should come next. One of these studies with the broadest approach, approached respondents in six countries – Argentina, Denmark, France, Japan, and the US. Among those surveyed, awareness was high, but frequent actual use was low. On a daily basis, frequent users ranged from 7% in the U.S. to 1% in Japan. 56% of 18-24 year olds had tried an AI service and 16% of those over 55.

My personal interest concerns AI in schools so I tried to locate studies that attempted to establish typical patterns of use by secondary students. Here is a 2024 study from Commonsense Media on this topic available to all online. A very short summary concluded that will half of 14-22-year-olds have used an AI service, but only 4% report being daily users. Beyond these basic statistics, I found it startling that minority youth (Blacks and Latinx) reported a higher frequency of use – 20% – 10% claimed to be weekly users. I cross-checked this result several times to make certain I understood it correctly. When asked to categorize their use, young people reported searching for information, generating ideas, and school work in that order. Another large category of use was generating pictures. The authors reported some concern when finding that searching for information was the most frequent category of use.

Participants were asked about concerns that limited their use of AI and potential accusations of cheating were high among these young people.

I admit I need to review this study more carefully because it is not clear to me if the participants were including any classroom use in contrast to what I would call personal use.

The “what can I do with this” question

Mollick and the 10-hour investment. I have read several efforts by Ethan Mollick (NYTimes, Kindle book) and find his perspective useful. He claims using AI is different from learning other technology applications in that there are not exact instructions you can follow to find productive uses. Instead, he proposes that you invest 10 hours and try the tool you select to accomplish various tasks that you face daily. If you write a lot of emails, chat with the AI tool about what you want to say and see what it generates. Request modifications to improve what is generated to suit your needs. Ask it to create an image you might have a way to use. Ask it to generate ideas for a task you want to accomplish. Some may tell you that AI is not a person and this is obviously the case, but forget this for a while and treat the AI service like an intern working with you. Converse in a natural way and give a clear description of what your tasks require. Ask the AI service to take on a persona and then explain your task. If you are trying to create something for a classroom situation, ask the service to act as an experienced teacher of XXX preparing for a lesson on YYY. Expect problems, but if you involve the tool in areas you understand, you should be able to identify what is incorrect and request improvements.

I watched the recent Apple announcement regarding the company’s soon to be released AI capabilities. Thinking about Apple’s approach, I could not help proposing that experiences with Apple products in the ways Apple plans could be a great gateway to finding personal practical applications of AI (Apple wants you to think of their approach as Apple Intelligence). Apple intends rolling out a two-tiered model – the AI capabilities available in a self-contained way on Apple devices and AI capabilities available off device. The device-located AI capabilities are designed to accomplish common tasks. Think of the on-device capabilities as similar to what Mollick proposes – ways to accomplish daily tasks (e.g., summarization, image creation, text evaluation and improvement, finding something I know I read recently). AI capabilities are available within most Apple products and also within other services. I could not help wondering how Grammarly will survive with AI tools available to Apple users who own recent Apple equipment.

Obviously, I have yet to try the new Apple Intelligence tools and I doubt I will close out my AI subscriptions, but I do think Apple tools as a transition will increase day-to-day usage.

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.