Month: February 2024

Posted on

Chat GPT: It’s all About the Prompts

Written by Dr. Charles R. Ciorba

To get the most out of ChatGPT, it is best to master the use of prompts. Doing so can really can make a difference

Starting a Chat

If you enter a basic prompt that is similar to something you enter using an internet search engine, you will get a simple response.  If you make this mistake, you might become bored and frustrated. As a result, you could very well assume that you can do better using the internet.  Basically, the rules for using key words in internet searches do not necessarily apply to ChatGPT.

According to Baker (2023), it is best if you use this form of AI differently.  For example, you can actually assign ChatGPT an identity.  Perhaps, several identities.  You can also assign yourself an identity.  Tell it to address an audience such as a (a) group of teachers, (b) group of fifth grade students, or (c) an angry parent.  Set the stage correctly by giving it a command, asking a question, or telling it to behave in a specific manner.  Ultimately, you are looking to receive a detailed and well-defined response. 

Here is a direct quote from the Baker text (2023, p. 49).

  • Who do you want ChatGPT to be?
  • Where, when, and what is the situation or circumstance you want ChatGPT’s response framed within?
  • Is the question you’re entering in the prompt the real question you want it to answer, or were you trying to ask something else?
  • Is the command you’re prompting complete enough for ChatGPT to draw sufficient context to give you a fuller, more complete, and richly nuanced response?
  • Is your prompt specific and detailed, or vague and meandering?  Either way, Chat GPT will mirror it in its response.

I mentioned this in the last learning module, but when you are finished with a specific conversation, it is always best to close out the conversation before starting a new one.  If you hold a long conversation that involves several topics, ChatGPT may get confused.  If this happens, it might get offensive or start giving you wrong answers.  I am not sure how this service can get offensive, but it may call you offensive names. I myself am not at a point where I want to intentionally make this happen.

Reviewing Your Chat History

On the left side of the ChatGPT screen, you will find a list of your most recent conversations. You can always return to earlier chats, which can be quite handy, but you only get so much storage. So, you can always delete individual chats and paste them into a word document.  If you hit the “Clear Conversations” tab, you will delete your entire chat history.  By doing so, your screen will be clean, but it does not delete the information from the open AI servers. As I mentioned in the previous learning module, be careful not to enter confidential information (passwords, finance information, etc.)

Understanding Prompt Engineering 

The term prompt engineering refers to formulating all prompts in natural language. I am told there are prompt forums on this internet, so feel free to check them out.  Here is one example: https://github.com/f/awesome-chatgpt-prompts. 

Token Limits and Storage Issues

ChatGPT keeps a record of all of your prompts, yet it can only remember 3,000 words from a conversation (4,000 tokens). This may explain why ChatGPT can get confused if you bring up too many topics in one conversation.  Prompts are broken down into tokens, which are otherwise known as “pieces of words” (p. 55).  Here is another direct quote from the Baker text (2023, p. 56):

  • 1 token equals about 4 characters.
  • 100 tokens translates to about 75 words.
  • Two sentences equal about 30 tokens.
  • A typical paragraph is about 100 tokens.
  • A 1500 word article totals around 2048 tokens

A single chat is limited to 4097 tokens.  In addition, if your prompt is long, it will be cut off at 97 tokens. A Tokenizer Tool can be found at the following link: https://platform.openai.com/tokenizer. Please keep in mind that as ChatGP develops, these limits can change.

Thinking in Threads

All of us probably know that a string of messages is referred to as a thread.  As such, it is best to write prompts as part of a thread as opposed to a single question.  By thinking in threads, you can control the direction of the conversation. Be forewarned: instead of asking for clarification, ChatGP will start guessing. 

Chaining Prompts, and Other Tips and Strategies (pp. 58-59)

Writing good prompts takes time.

This is where the true art lies when communicating with ChatGPT.  It doesn’t get easier, but you will improve with time.

Define your goal.

Determine exactly what you want from your conversation.  Craft your prompt to achieve your goal.

Be a storyteller.

Remember, you can give ChatGPT an identity. You can tell it to provide answers as a school administrator, a parent, or a musician.  If you give it a sample of your own writing, it will do its best to mirror your writing style.

Any undertaking is acceptable.

Ask ChatGPT to check your student’s homework to find errors.  Ask it to finish a thought or an exercise.  This is where ChatGPT can make the life of a teacher easier.

Be specific.

Include as many details as possible when writing your prompts but be aware of your token limits.

Strategize

Tell ChatGPT about the audience you are dealing with.  If you say you will be working with a group of elementary students, ChatGPT should be able to formulate the language to address that population.

Reference

Baker, P. (2023). ChatGPT for dummies. Wiley.

Posted on

ChatGPT: Resistance is Futile!

Written by Dr. Charles R. Ciorba

ChatGPT: What is it?  How does it affect me as an educator?  As of recently, I was unable to answer either question.  I first heard about this new phenomenon last spring when I was driving home from a late-night practice session on campus. When driving at night, I usually listen to WPFW, which is a public radio station in Washington, DC. Their playlists are fantastic. Anyway, the DJ on the air was talking about ChatGPT. He said it was going to be a “real game changer.” So, I was made aware of it, and I held off learning more about it until now.  After all, I’m currently teaching a class that centers on the future of music education.  Can ChatGPT be a useful tool for us?  Let’s find out.

What is ChatGPT?

ChatGPT is referred to as a large language model (LLM). It is a representive generative AI. This means it can actually create new content, as opposed to analyzing data that already exists (e.g., Siri).  You can actually start a conversation and ChatGPT will generate a discussion. If this is enough to stimulate your curiosity, simply go to https://chat.openai.com/ and create an account.  It’s a very easy process but be forewarned: the information you input (and the ensuing conversation) is not private, so be careful not to disclose any confidential information.

How it differs from what we’ve used in the past.

Once again, ChatGPT is a generative AI model. This is different from some of the AI tools you may have used in the past.  For example, if you want to know where the good vegetarian restaurants are in Northern VA, Google Maps uses a ranking system based on the information you enter. The information is usually accurate, and you can choose a location easily.

ChatGPT is different. You can actually engage in a dialogue, as it is a system that can generate its own content. This sounds absolutely fascinating, although the accuracy of the generated content is not always consistent. It generates content based on patterns it has learned.  Sometimes it delivers the truth, and sometimes it doesn’t. It can lie, it can hallucinate, and like humans, the responses can sometimes be inappropriate. It is not the fault of the machine, as it does not think like humans. Rather, it responds based on pattern recognition in 95 languages.    

Fears of ChatGPT

What a fascinating technology, but it has generated certain fears among the public.  A number of lawsuits have been created against ChatGPT, mostly surrounding copyright.  Currently in the United States, AI generated content is not protected by copyright law.  In addition, there is a gray area when examining who is liable for incorrect information generated by ChatGPT. I for one would not use it to enquire about stock information or medical diagnoses.

Of course, one needs to be careful of privacy issues, biases, deep fakes, etc. We also need to watch out for reality apathy, which occurs when the public doesn’t know what is true, but we don’t want to engage in the effort to determine the correct answer. 

Is ChatGPT coming after your job?

Now is not the time to panic about machine overlords.  That type of AI is known as artificial general intelligence (AGI).  We may have to deal with it someday, but for now, it’s not something to worry about.  Can ChatGPT replace teachers?  I don’t think so, and I believe it can make education more efficient. It can also help to eliminate some of the frustrations we deal with as teachers.  As for replacing us, humans have an advantage in that we have the ability to use our creativity and intuitive intelligence. In addition, our brains are very efficient, while AI consumes massive amounts of energy. It is important to keep in mind that ChatGPT has a great deal of information, but that is not the same as knowledge. As for now, humans have the upper hand. 

Using ChatGPT

As teachers, we may wonder if students are using ChatGPT to complete their homework.  It seems Chat GPT is not very good at solving riddles or algebraic story problems. It can also give erroneous answers without warning.

I have been told that it can generate papers with perfect English and punctuation, but from my perspective, this would send up a warning flag.  I’ve graded many papers and supervised many dissertations.  If someone turned in a paper written in a perfect manner, I would be suspicious.  Then again, as music teachers, I wonder how we can use this fascinating tool in a useful manner.

As for using ChatGPT, I created an account, and as soon as I logged in for the first time, I created a prompt asking what notes make up a Coltrane pentatonic scale starting on E.  Here is the response:

The Coltrane Pentatonic Scale, also known as the Coltrane Matrix or simply the Coltrane Scale, is a symmetrical pentatonic scale created by the jazz saxophonist John Coltrane. It’s derived from the circle of fifths and consists of major thirds and minor thirds.

Starting on E, the Coltrane Pentatonic Scale would consist of the following notes:

E – G♭ – A♭ – B – D

This scale can be seen as a series of major and minor thirds, giving it a distinct sound that is often used in jazz improvisation.

The correct answer is E-F#-G-B-C# (1-2-b3-5-6).  So, ChatGPT provided the wrong answer, and I found the text description to be quite general.

I then asked ChatGPT to tell me what notes make up a Coltrane pentatonic scale starting on D.  Here is the response:

Starting on D, the Coltrane Pentatonic Scale consists of the following notes:

D – F♭ (E) – G♭ (F) – A♭ (G) – B (C♭)

This scale follows the same pattern of major thirds and minor thirds as the Coltrane Pentatonic Scale starting on E, but shifted to begin on D.

Wow.  This is not a correct example of the Coltrane pentatonic, and the enharmonic spellings for the third and fourth notes of the scale are incorrect. I think we can all agree that Gb and F are two different notes separated by a half step.  Furthermore, the intervals differ when comparing both examples provided by ChatGPT.

I then asked ChatGPT, “What are the Detroit Lions chances of winning the 2025 Superbowl?”  ChatGPT refused to provide me with a specific response.

I then asked Chat GPT to provide an explanation of Suzanne Langer’s philosophy of music as metaphor.  I did receive a bland and somewhat general explanation.

My first encounter with Chat GPT was less than successful.  Of course, I was using it much like I would use Google or Siri.  I am told success depends on how well you write your prompt.  As such, I blame myself for the inaccurate results, and I am committed to learning how to use Chat GPT in an effective manner.  With intelligence and creativity, ChatGPT can be a valuable resource.

After all, resistance is futile.  Much like radio, television, computers, the internet, cell phones, and social media, I believe ChatGPT is here to stay. There are many bugs to work out, but the potential for it to be a real game changer is undeniable. I look forward to learning more about this wonderful technology.  More to come!

Reference

Baker, P. (2023). ChatGPT for dummies. Wiley