Artificial intelligence. AI. It’s a gigantic topic, rapidly growing and advancing faster than most of us can comprehend. ChatGPT, ChatGPT-4, Microsoft Bing, and Google Bard are just the first AI programs that come to mind. ChatGPT was released last November and before I could even look into it, ChatGPT-4 followed in March. Short of being a developer in the business, how can anyone really expect to keep up with such rapid advances?
As it happens, my son is a developer, an IT architect, and he has a good idea of the pros and cons of this stuff. We’ve had some interesting conversations, particularly because my granddaughter is in high school. AI is a goldmine of potential for students — to write their papers, research their assignments, suggest ideas, do their math, etc., for their current classes and for college admission applications, job applications, etc.
Unfortunately there’s no surefire way, no test, for teachers, educators, admissions officers, or anyone, to detect AI-generated material. The best my son and I could come up with was totally unrealistic: Require all assignments to be done in class under teacher supervision after collecting all electronics at the beginning of each class. I even asked ChatGPT if there was a sure way to detect its use by a student and the answer was, essentially, no. Everything would depend on the student being committed to not using AI.
You can try ChatGPT for free. Just head over to the website and sign up. It’s interesting to play with and you might even find a good use for it.
ChatGPT has already scanned all the information on the internet or, in effect, all of human knowledge. From there it can deduce patterns of human thought and behavior — how we write, paint, build, compose, socialize, communicate, ad nauseam. Basically, it can learn how we think and emulate those patterns .
AI can also use its knowledge to create images in programs like Midjourney and DALL-E 2. You may remember last year when Jason Allen won the Colorado State Fair’s contest for emerging digital artists. His entry was titled “Theatre D’opera Spatial” and, as he made clear from the beginning, it was created using Midjourney. Artists across the country were outraged and accused him of, in effect, cheating. Allen contended that AI was merely a tool and he was the creator. (You can try Midjourney for free. Or at least it was free when I played with it a couple of months ago. DALL-E is still free to play with.)
“60 Minutes” recently did a segment on AI that I found enlightening:
AI is a HUGE, mind-boggling topic making news almost daily. (This week’s writers’ strike, for example.) So huge, in fact, that Geoffrey Hinton, the so-called “Godfather of AI,” left his position at Google last week. He tweeted, “I left so that I could talk about the dangers of AI without considering how this impacts Google.” AI, it seems, is advancing far faster than he thought it would.
“It is hard to see how you can prevent the bad actors from using it for bad things,” Hinton told the New York Times in an interview published Monday. “I console myself with the normal excuse: If I hadn’t done it, somebody else would have.”
“The dangers of AI.” If the godfather of AI is concerned about its inherent dangers, we probably should be too.
_____________________________________________________
Addendum: One of my ChatGPT queries
SusanR: How can we keep students from using ChatGPT to do their assignments?
ChatGPT: As an AI language model, ChatGPT is a tool that can assist students with their assignments and provide them with relevant information. However, if you want to prevent students from using ChatGPT to complete their assignments entirely, there are a few measures you can take:
- Set clear expectations: Let your students know that using an AI tool such as ChatGPT to complete their assignments is not allowed. Explain why using such tools is not acceptable and what the consequences will be if they are caught.
- Provide alternative resources: Make sure you provide your students with alternative resources they can use to complete their assignments. This may include textbooks, reference materials, and other reliable sources of information.
- Change the assignment format: Consider changing the format of the assignment so that it is less likely to be completed with the help of an AI tool. For example, you can ask students to write an essay on a particular topic instead of providing them with a set of questions that can be easily answered with the help of an AI tool.
- Monitor student work: Keep an eye on the work that your students are submitting and look for signs that they may have used an AI tool. This could include unusual language use or references to information that may not be readily available.
- Use plagiarism detection software: You can use plagiarism detection software to check student work for any signs of plagiarism. These tools can help you identify any instances where students have used AI tools to complete their assignments.
Ultimately, it is important to remember that the goal of assigning homework and assignments is to help students learn and develop their skills. If you are concerned that students are using ChatGPT to complete their assignments, it may be worth considering how you can modify your assignments and teaching approach to better support their learning.
(The obvious problem with #3 is that students will just ask ChatGPT to write the essay.)
Well, he’s right: somebody would’ve .. :\
But maybe not as fast or as well?
yes, when I saw he left for those reasons, my fear factor went up. I teach young children so not directly impacted by this, but for high school and above, what about having students orally share whatever they studied or wrote about to see what knowledge they really have gained, which is part of the goal, anyway?
That was one of the approaches the program suggested. It would require all parties taking the time and personal responsibility to do that. A willingness to give up all the electronic crutches. I was going to include what ChatGPT told me, but it was lengthy and I thought no one would want to read it all.
Changed my mind and added one response from ChatGPT.
This is a fascinating topic. I can’t help but feel that AI could change civilization. And what do we do with all the time we have when AI can run our lives?
The possibilities are limited only by our imaginations and changes are already occurring. And now, for better or worse, we have AI to help us come up with new ideas even faster than before.
There are many people already fearing they will lose their jobs to AI and have nothing but time on their hands.
Perhaps all school essays should be written in class without access to AI? This could be the end of the term paper in education.
Doing everything in class without electronics was the best my son and I could come up with. But that would be a huge time sink for both students and teachers. Term papers may still exist, but I’ll bet none will be done without AI. If there’s an easier, faster way to do the assignment, whatever it is, kids will use it.
At the small college where I work, nursing students take their exams in a “lockdown browser” app that doesn’t allow them to use the computer for anything but the exam, and they have to turn their cell phones off. The “learning management systems” are rigged with plagiarism detection sofware. Higher ed has already acclimated to e-threats to academic integrity.
That said, I’m still fairly worried about AI, mostly because the open source orgs like OpenAI make their code available to everyone, including the criminally motivated. And then you have nations like Russia, China and North Korea, all of whom are bent on besting the USA. Not a healthy scenario.
It’s good to know higher education is keeping up with this stuff. I should have guessed they’d be well ahead of my awareness and concern. I don’t know where the high schools are on this but I hope they are staying abreast of AI developments (while being distracted by more immediate concerns like gun violence).
I hope Geoffrey Hinton will be joined by others like him who understand the technology and its dangers. He said he’d thought the potential we’re seeing now was 30-50 years away! I’m just afraid the horse is already out of the barn.
Gee, you don’t suppose it was profit motive that prompted OpenAI and others to release all this to the public???
Scary stuff. The only thing I can think of to compare this to is atomic energy.
Hmm, I’d not thought of that, but it seems like a fair comparison. So many good things about atomic energy, so much it can do for us … along with the ability to obliterate us if used with carelessness or evil intent. The more I think about it, the better the comparison seems.
Thanks for sharing. I spent quite some time on trying to dispel the rabbit hole and here you can find a few pointers for taking AI a piece by piece
Introduction to generative AI
https://francescolelli.info/machine-learning/an-introduction-to-generative-ai/
Introduction to Neural Network:
https://francescolelli.info/tutorial/neural-networks-a-collection-of-youtube-videos-for-learning-the-basics/
Introduction to Knowledge graph
https://francescolelli.info/big-data/on-knowledge-graph-and-artificial-intelligence/
Introduction to Genetic algorithms:
https://francescolelli.info/machine-learning/on-genetic-algorithms-as-an-optimization-technique-for-neural-networks/
A collection of free resources on AI:
https://francescolelli.info/uncategorized/a-selected-list-of-free-material-for-learning-artificial-intelligence-ai/
A collection of courses on datacamp that prepare you on data science skills:
https://francescolelli.info/uncategorized/a-selected-list-of-free-material-for-learning-artificial-intelligence-ai/
hope it can help! your feedback is very welcome!
Thank you, Francesco. Clearly you’ve spent much more time delving into AI than I have or likely ever will.