What about AI?

Today we are going to talk about Pi Ai. Pi is a free conversational AI. It is a terrific tool for troubleshooting, brainstorming, or preparing for a conversation you are not sure how to proceed it.  Think about how many of our students need help practicing for job interviews. 

I played with it a little bit.  For the most part, I liked it and I can see how the process works.  You can load the Call Pi app in iOS or Android and speak directly with the Pi.  Just ask it a question and it answers.  As a test drive, I asked it to ask me five questions common in a job interview for a librarian at a community college and evaluate the answers.  It immediately started listing all five questions so I told it to stop.  You actually have to push a button to interrupt.  Then I told it to ask a question, wait for an answer, evaluate and comment on that answer, and then move on to the next question.  It did exactly that.  The responses were great.  I answered questions like I might answer them in an interview.  It basically said that was great but you could also include, x, y, and z.  It was right.

When you are finished with your conversation, a transcript is right on the screen so if you did get helpful advice you can remember it.  I can see how this would be really helpful.  Some of the things it does that are advertised on the log in page are journal keeping, making plans, planning a vacation, reading stories, or learning new things.  It is definitely worth your time to explore it.

I have read several articles lately that discuss returning to the blue book as a means of keeping students from cheating on tests.  That is one option, and it may not be a bad one under certain circumstances.

If work is done in the classroom with no devices the students have to demonstrate their knowledge. Of course, that would mean you can’t use class time for teaching, so there is definitely a sacrifice if that is the route you choose to go. 

Why not turn it around? Instead of homework, their hours outside the classroom are spent preparing for the upcoming class.  We know they are not doing this now, but once in-class performance and grades are tied to it in a very tangible way, it may improve.

Another option is to provide specific guidance in the use of AI, even requiring a specific AI program and identifying which things they are allowed to ask for from it.  Require them to include their prompts as part of the assignment.  You will have to test the assignments yourself to know what responses they are likely to get from the different AIs so you can pick the right option.

Neither of these options reduce your workload, but they both increase student knowledge.  The second prepares them for the future where AI is a routine thing.

Still referring to the Chronicle of Higher Education article, The Student Brain on AI, there is one other thing to consider.  Researchers say that one of the most significant — and troubling — findings to them was that the essays written with the aid of ChatGPT showed a greater degree of homogeneity.”  Being able to regurgitate facts is not really a goal of higher education.  We want to teach them to think.  That involves looking at issues from various points of view and developing opinions.  You could always include as part of your assignment that they have to provide a statement of stance.

One thing we have to consider is that people will always opt for convenience, especially under stress. Most students think they are under stress (just wait until they face the real world, right?)  Even professors take short cuts.  The problem isn’t really AI, it is the wrong use of AI.  That is what this blog is all about – finding good uses for AI. 

You could consider an AI Tutor, like Tutor AI. As a teacher, you can create a course around your own content and the AI provides the tools students need to understand the assignment.  The AI program can create flashcards, tutorials, exercises, and more.  It is probably worth a few minutes of your time to see what it can do.

There was a lot of great information in last week’s Chronicle of Higher Education article called The Student Brain on AI.  One important point it makes is that change is hard.

“New technology has always brought with it social panic. Commentators argued that the telephone would be ruinous to families because women would spend more time gossiping.”

And another one:  “When the internet and search engines began to be an integral part of life, people worried about their effect on our memory and intelligence. As a viral 2008 article put it, “Is Google Making Us Stupid?” “  Well, I will give that one a little credit.  Google has ruined our ability to identify false or bad information.  Then again, information literacy is a topic that can be taught.

One important facet was explained by Derek Bruff, associate director of the Center for Teaching Excellence at the University of Virginia, wrote about the MIT study. “When we ask students to do things that are neither interesting nor relevant to their personal or professional lives, they look for shortcuts.” That has always been true.  Today the tool is different.

BLUF:  Probably not.  At least no more than any other technology.

For those of you who are unaware, the library subscribes to The Chronicle of Higher Education.  Please come by to read it sometime.  This week, one of the headline articles is called The Student Brain on AI and the results are not nearly as bad as rumors and social media would have us believe.  Let’s start with their assumptions:  “Students use artificial-intelligence tools regularly to tutor themselves, organize their thoughts, spark ideas, and edit their writing. Is that weakening their ability to think?”

On the upside, I hope that students are using these terrific tools for exactly what the article states: tutoring, organizing, idea generation, and editing.  I think those are the best uses.  We run into problems when students use them to do the work for them.  Last night I had a student in the library struggling to choose a topic.  The “world issue” umbrella he was given was just too big.  I sat him down and we put the assignment parameters into Gemini and asked it to provide a list of potential topics.  It gave us about a dozen ideas.  The student picked one and we started the appropriate research.  Simple and useful.  Personally, I don’t see that this rots a brain.  Rather, it allows us to move past the drudgery and into the actual learning part.

The other important idea from the article is that journal articles about the effect of AI on the brain cannot be definitive.  The target is moving too fast.  “AI is now embedded in so many products that it’s impossible to isolate how it is shaping us.  And, perhaps most significantly, learning is a complex process that is hard to define, let alone measure.  Trying to determine the interplay between our use of technology and our own thinking is tricky.”

One thing is for sure: If an article about whether AI is good or bad is more than a minute old, it is probably out of date.  Even this one.

I think AI can be a bit overwhelming.  That is why I started this blog.  I also think I have pretty average tech skills.  The parts I know, I know really well, but mostly I know what can be done, not necessarily how to do it. I am also really good at watching YouTube tutorials.  A lot of us fall into that category.  

Well, AI is another tool in that toolbox.  The general, generative AIs, like Gemini or ChatGPT, are super easy to use and the responses are almost conversational.  When you are struggling to get Google to understand what you are asking for, try an AI.  Small bites are always good.  Instead of doing a bunch of research for my Heritage Trail windows, I asked it to name two fun things to do in every county in the Pecos Trail region.  It did exactly that.  If you want to know what they were, read VP Arca's weekly newsletter for my More than Just Books articles. 

These little things are a great way to familiarize yourself with using AI and learning all the ways they can help, without making you feel like you are cheating.

Regular readers know that I think Consensus is a great tool for college students to be guided to for research.  Recently, they added a search history page for those who are logged in when they do their searches.  I don’t know how organized you are, but I find myself doing the same searches multiple times when I am working on a big project.

For our Allied Health department, Consensus searches now support MeSh synonyms to improve the accuracy of medical queries.  Remember, all of PubMed is included in Consensus, plus many other related resources.  Now you can get it all in one place.

As most of my regular readers know, I read widely on AI subjects and very often find things I want to share.  This week's post comes from David Moldawer’s blog, The Maven Game, and his post, The Power of a Blue Book.  Very much of what he said is what I wish I had thought of myself, but like the cheating mentioned in his post, I am letting someone else do my thinking for me here. 

What we have to recognize is that students are going to use AI, they are going to use it to cheat, and the detectors we use provide so many false positives that they are almost useless. The battle against LLM abuse by students is already lost.  So what can we do instead? 

First, acknowledge the long-known fact that writing is a significant part of learning.  Creating a sentence is hard work and creating multiple sentences that go together coherently is really hard.  If we let LLMs do the writing, we are doing a disservice to them and us. 

Second, we have to acknowledge that the only solution is to write.  In the classroom, during class time, create assignments that require them to think and put it on paper.  This isn’t easy, I know.  We already have too much to cover and too little time to do it, but we also sacrifice hard assignments for simpler things that can be easily graded by Canvas.  Students need hard assignments to learn and we have to grade them. 

I thought I was doing this for my History 2321/2 classes at Dallas College.  The plan was to spend the first ten minutes of each class having the students write a summary of the readings they were supposed to have done before class.  I was a new teacher and let it go a little awry.  They would come to class with their book, reading for highlights and copying them down. At first, I let them, thinking that at least they were looking at the book.  Soon, though, I had allowed a pattern to be established and I did not know how to break it.  Grading wasn’t much work.  There was never more than half a page.  I could look at it for a few seconds and catch anything wrong.  It was pretty much a matter of ticking off the assignments as complete.  Now, if I had not allowed books, they would have had to think for themselves, they would have gotten in the habit of doing the reading before class, and the grading would not have been any harder.  Yeah, I will do that next time.  You could try it.

Consensus was great when it provided abstract information for relevant journal articles, but they have updated their access, and now you get search results that pull straight from the full text of the papers.  In some ways, this will be great.  We all know abstracts rarely provide the information you want.  With the new full-text access, you won’t have to search through a 120 page article to find the relevant parts.  Consensus does that for you.  Of course, it is only a program and is working from your prompts, so some tweaking will definitely be required.  It is also a good idea to pull the full articles to make sure the AI did not miss anything, but this is another tool that will save you enough time to make it worthwhile.

Data Analysis and Prediction

On the surface, these may not seem like student tasks, but analyzing statistics and being able to make informed choices based on your findings is exactly what students should be able to do.

• Tasks: Extracting insights from data, identifying patterns, and forecasting future trends.
  • Predictive Analytics:  Forecasting future events (e.g., customer churn, sales forecasting, fraud detection).
  • Data Cleaning and Preparation:  Automating the process of making data ready for analysis.
  • Anomaly Detection:  Identifying unusual patterns or outliers in data (e.g., fraudulent transactions).
  • Business Intelligence (BI) Augmentation:  Enhancing traditional BI with AI-driven insights.
  • Risk Modeling:  Assessing and quantifying financial or operational risks.

• Programs/Platforms:
  • Google Cloud AI Platform (Vertex AI, BigQuery ML): Comprehensive suite for building, deploying, and managing ML models, including predictive analytics.  These are really big picture programs for creating entirely new AI programs.       
  • AWS SageMaker:  Fully managed service for building, training, and deploying machine learning models.
  • Azure Machine Learning: Microsoft's cloud-based platform for ML.
  • TensorFlow & PyTorch:  Underlying frameworks for building custom predictive models.
  • R:  Statistical programming language widely used for data analysis and machine learning.
  • Databricks:  Data platform that integrates AI/ML capabilities for large-scale data analysis.