The Ultimate Guide to ChatGPT chat with Belinda and Anthony

Welcome to your mega file for the recent ChatGPT event with Belinda Brooks and Anthony Snape, authors of the Ultimate Guide to ChatGPT.

In this blog, you will be able to access:
* The recording of the Zoom event
* The transcript created of the ChatGPT event by ChatGPT (any more meta and I would cry, right?)

Some casual observations on using ChatGPT for transcription work

It knew the Aboriginal and Torres Strait Islander country right off the bat
It corrected snafus that Otter had made in the transcript using context
it really does suck at names, even if you identify them first (sorry Belinda!)
Questions are key – but not even consistency in what you ask will get you through
ChatGPT began answering some questions about itself instead of tidying the references
It also made assumptions about Belinda’s career based on information from Anthony and assumed the two write songs together when this is not the case.
It also skirted over references like Nazi (in context, I asked about the hacking of a chat bot years ago that ended up spouting right wing propaganda before devolving into Nazi hate speech before being pulled off line.
It also responded later to a question of me asking what it’s nickname was. At first, it claimed it was AI and that wasn’t relevant. Then out of the blue on a different day, it circled back to that question and answered “You can call me Chat if you like” on a different day but the same thread. Eep!
I melted ChatGPT several times by asking it to write about itself. This may be due to it’s desire to answer the questions. Or it might be the quality of the transcript included too much human nuance together with hinkey translations that it had to battle. Or maybe I just melted the robot within!
It tends to tire. The more I pushed on threads, the more ChatGPT’s focus wandered and varied.
Our jobs are safe. Not only do AI tools grab the content, but they also grab the bloat, too. You need to edit the work more than you otherwise might.

TRANSCRIPT DONE BY CHATGPT

In this interview, we have Belinda Brooks and Anthony Snape who will give us an inside look at the old ChatGPT. As a chat-based product, there has been a lot of talk about it lately. Belinda and Anthony wrote a book about ChatGPT, which is fascinating since they are an End of Life Consultant and a musician, respectively. If you need to leave, don’t worry; the link to the transcript will be available afterward.

Before we begin, I’d like to acknowledge that the land we are on belongs to the Dharawal nation of the Wodi Wodi people, who have fished, prawned, and loved this land for thousands of years. Our area has at least two Dreamtime legends, the formation of Lake Illawarra and the story of the five Island sisters. We pay our respects to any Elders past, present, and emerging and thank anyone with Aboriginal or Torres Strait Islander heritage for joining us today.

Belinda and Anthony will discuss ChatGPT, and we ask that you keep your mics off to allow them to have a clear run as they chat. Please feel free to ask questions in the chat, and there’s an added incentive as Anthony will use ChatGPT to write lyrics based on your questions, and he will make a song out of them if time permits.

Now, without further ado, let’s hear from Belinda and Anthony as they introduce themselves.

Anthony Snape: Hi, I’m Anthony Snape, a singer-songwriter. I’ve always been interested in technology and its use in music. I became interested in chat GPT through Dall-E, an AI image creator that did beautiful and in-depth work before the rise of current AI software. During COVID, while playing for my online audience, I got interested in applying technology and came across TPT in its early development stages. At first, I was worried about the impact of this technology on our lives, but I pushed through to the other side and now look for ways to use it to get ahead in my career.

Belinda Brooks: Hi, I’m Belinda Brooks, an entrepreneur and member of Freelance Jungle. I’ve worked in community development, the end-of-life space, and as a director of people and culture for Soldier On in Canberra. I’m also an artist and singer, and I met Anthony four months ago and have been helping him with his marketing and promotions. We became interested in chat GPT to progress Anthony’s career further. We wrote a book about it to understand how it would affect our respective areas of music, arts, and freelance consulting.

What is ChatGPT?

For those unfamiliar, GPT-3 stands for “conversational, generative, pre-trained transformer.” Essentially, it is an AI tool designed to create human-like language and interactions by generating text. However, it’s important to note that it’s not a magic solution to all problems – rather, it uses statistics to predict the most likely next word in a sequence based on the data it has been trained on. Currently, it has access to a corpus of data drawn from various sources such as books and documents, but it hasn’t crawled the entire internet yet. Its programmers give it the information it needs to generate responses to questions, and it works to predict the most likely answer based on that data.

Overall, GPT-3 is a powerful tool for generating text and conversations, but it’s important to understand its limitations and how it works. As advancements continue to be made in AI, it will be interesting to see how GPT-3 and other language models evolve and improve.

Bek: For context, if someone says, “I’m going to write a poem about penguins,” it will think, “Okay, I need to go into poetry phrasing. And here’s some information about penguins that would suit.”

Anthony: That’s pretty much it. And it’s based on data that it had been given until 2021. So for the sake of today’s talk, I think we’ll just stay on GPT 3.5. We’ll be talking about the free version. And if we talk about a free version that everyone can access straight away or live, there’s paid versions, and just in the last 24 hours, they released a whole series of plugins that just blew my mind this morning. Like it’s ridiculous because every day, there are changes happening so rapidly that can completely change how you operate and do the day-to-day things in a business sense. It’s really exciting, but it’s also very overwhelming because it means that for that, there are plugins they are offering from yesterday, it means that you can automate a lot of the chat activity inquiry and also allow it to go out into the world and to do things like purchase orders and things on its own when you’re giving it a certain amount of parameters.

So basically, little tiny AI bots operating for your business autonomously, which is absolutely mind-blowing, and the potential for what you can build as a single person, working through that collection of opportunities is crazy and very overwhelming. And it takes a little bit of a learning curve to get through that. Just from the information this morning, it sent me down that path. I’m like, “Oh, my goodness, I’ve got to spend the day on this now and to figure out exactly what apps are available and what it can do.”

Bek: Because that’s what every freelancer needs, right? Another tool that completely overwhelms them and gives them lots and lots of work. Yeah. How are people using it in business? Like, how are our clients using it? Or how are the freelancers that you’re talking to applying GPT to their business and to good effect?

Anthony: Initially, most people started using this technology as conversational agents, as those chatbots that we’ve sort of seen and been able to use by, again, giving it another corpus of data, normally in relation to the business itself and how the business runs about its products. And then utilizing that as a customer service agent being able to answer those questions more consistently, possibly, than a team of customer service agents that are sitting there in the background. But again, there are limitations to this.

Belinda: You’ve probably seen in the media more recently, there’s been a lot of people who’ve been using this to generate the drafts for text and then editing them or failing to edit them, which is really problematic. Because we were working on something recently, and Anthony’s done a lot of crowdfunding when it comes to his music. And I wanted to see what changes. And again, that was prior to 2021 and current. And I wanted to see what it was going to bring up and then what it was going to give me when I specifically asked, “Okay, tell me about this man, Anthony Snape, and what he’s done in this space.” And what happened?

Anthony: Well, it came up with a lot of really great information that was sexual. And then at one point or another, it leapt into the totally creative and made up a whole ton of things that apparently I had done and was created that just was totally fictitious, including a new album, which the title was actually quite in alignment with something that he was working on.

An amalgamation of Anthony and Bek: An interesting thing that I heard the other day is that the initial version of GPT-3.5, before the current free version that you can access, had a factual rate of around 10%. However, when there’s a free version that everybody can access, there’s also a thing called the playground where you can create chatbots for your businesses. This is a paid service and you pay by collecting tokens. It covers many things and is difficult to put in a box.

Belinda: Translation is another area that people are really using GPT-3 for, and it works really well. It has been trained on books across multiple languages, making its ability to translate from one language to another very good. I was in India in December, and I had a local Bengali woman send me a beautiful folk song. When I tried to search the internet to find a translation for the song, I found nothing. So I ran it through GPT-3, and a friend of mine who was there confirmed that it was a perfect translation, except for the missing divine aspect of the original song. This was the first-ever translation that likely existed on the internet for that song.

Bek: To think about these sorts of things is pretty cool because we’ve all struggled with Google Translate and had someone laugh at us in Spanish. But what I’m hearing is that GPT-3 is a tool, and we have to have oversight as human beings. We have to go through and fact-check and make sure that things are on the level. Before you post anything that you put your name to, it’s definitely a good idea to go ahead and edit it. Editing operates really well with GPT-3 in adjusting formats and removing lines.

Belinda: Before you post anything, it’s a good idea to put your name on it. Editing is a great tool, you can adjust formats, remove lines, and make text easier to read. With TPT, you can even ask it to reinterpret data for an eight-year-old or write a play based on the information you provide. The possibilities are unlimited.

Bek: That’s fantastic.

Belinda: In relation to my book, I had a rough idea of some of the things I wanted to cover, but I asked ChatGPT to give me a list of 20 things to research. Then I did some off-ChatGPT research and asked it to suggest chapters based on my interests. It ended up writing a significant part of the book itself, but we had to edit and make adjustments along the way to make the content more engaging and flow better.

Anthony: Yeah, we noticed it was a bit repetitive, but we wanted to use it as an exercise to capture a time capsule.

Belinda: It’s like Harry Potter, but classy. Do we have to add an eye of newt to this one, or do we have to add something real?

Bek: I had met a tattoo artist years ago called Happy who used to wear jeans all the time because he practiced tattooing on his left leg before he went and worked on customers. So it sounds like the good thing to do is to actually play with it with your own content marketing, play with it with your own social media and your customer service, chatbots and all the rest of it, so that you can learn how to direct it better for your clients.

Belinda to Anthony: And you were talking yesterday about using it as a marketing assistant or just a general office assistant. So do you want to talk about that?

Anthony: Well, yeah, I mean, you can use it in so many different ways. And it all depends on what your particular business is, you know, maybe to automate some simple tasks that you do from day to day, even if it’s just moving data around spreadsheets and things like that. There are tons of different companies that now link up with ChatGPT. And there are a whole bunch of plugins that are available natively now on Chrome, where they incorporate ChatGPT and also give you some hints and stuff about prompt engineering. So you can be really specific about the ways that you use this. A simple way of using it is just within emails, allowing it to, there’s now inside of Gmail, either a plugin or something, but basically, you can just get it to reply to your email, and then you can read over it and just send it off. It means you don’t have to spend a lot of time coming up with intense emails. Also, we tried this thing the other day, where it’s like, okay, so what happens if you’re having a wonderful… (transcription trails off).

Belinda: So this is what we weren’t in and this was in the free version. I will delete the swear words that we added in because we deliberately added swear words in there just to see what was going to do with it. Yeah, so we’ll go blank, and you just figure out what those blanks are. Okay. But I just knowing some of the conversations that happen. On Freelance Jungle, I thought this might be useful. I just wanted to try it.

I’ve just had this terrible situation where a client has gotten back to me, and he’s been really irrational and doesn’t understand the products that I provide. Even though I’ve explained it thoroughly a number of times, I want to tell them to blank. Blank all, but I want to tell them in a professional sort of way. Could you please help me write something that achieves this outcome?

ChatGPT: I understand that you’re really frustrated with the situation. But as a professional, it’s important to maintain a level of decorum in all communications, especially when dealing with clients. Instead of using derogatory language, or telling the client to what I mentioned before, I suggest taking a more diplomatic approach to address their concerns. Here’s an example response that you could use: “Dear client, I’m sorry to hear that you’re feeling unsatisfied with the products I’ve provided. I’ve done my best to explain the features and benefits of our offerings, but it seems that there is still some confusion on your end. I would be happy to set up a call or a meeting to further discuss your concerns and see how we can find a solution that works for both of us.

Alternatively, if you prefer, I can provide additional documentation or resources that may help clarify any questions you have a value of our working relationship. And once we ensure that we find a way to move forward in a mutually beneficial way, please let me know how you’d like to proceed.

Belinda: Especially in those moments when I’m frustrated, it’s hard sometimes to write the initial response that you want to send to a client, knowing that it won’t be appropriate. I know that before I even start because there are still a lot of issues that need to be addressed. However, having somewhere to start is helpful. There are a lot of things in there that I wouldn’t necessarily use if I was writing the response, but it’s giving me a solid starting point.

Anthony: And the reason that the conversational aspect of ChatGPT is important is that I always start my conversations with “Hi.” This reminds me that I am having a conversation and that I am explaining to the program who I am, where I’m coming from, and what kind of outcome I want. The more information you provide in a chat situation, the better the responses will be.

Belinda: As a freelancer, I think that’s one of the ways we can utilize these tools. You can say to ChatGPT, “I’m a freelancer, this is the area of work that I’m involved in. These are who my clients are, this is my avatar. How can you help me?”

It’ll come back with a whole list of things that it can do to help you. For example, it might say, “I can help you with your marketing.” You can then ask it specific questions like, “How can you help me with my marketing?” or ask it to create a marketing plan for you in relation to your avatar and your personal business. You can keep asking questions and even ask the technology to keep asking you questions until it has enough information to give you a thorough response.

The conversational aspect is important and that’s what it was designed for initially. We can utilize it as an assistant in our own businesses if we let it. I even asked it the other day to write some points for me to add to advance care directives that would meet client needs and work in a way that would be suitable for doctors and nurses. The quality of the response you get is related to the quality of the input that you give it to provide that response.

Bek: Absolutely. And it looks like you hit the nail on the head with the freelance stuff as well. It takes the emotion out of the situation, so it kind of acts as a little buffer there. Like you said before, Belinda, it’s almost like mommy’s teaching you how to say it better and then giving you the opportunity to say it better. And I do love that whole idea of speaking the duality across legalese or medical stuff or whatever.

Bek: There is a lot of stuff about freaking out about whether it’s going to steal our jobs. Do you think that that’s going to happen anytime soon? I’ll take that from Anthony, because he initially had those feelings.

Anthony: I think the timeframe is the question. Absolutely, it’s going to take all of our jobs, no question about it. It’s already happening. Don’t freak out right now. As to what elements humanity can provide that are valuable, we’re still learning. We’re still learning, and I think we’re going to be looking inward and outward, going, “Okay, what is it that I bring to the table that is useful and interesting and that can help my fellow humans?”

That is some of the really big questions we’re going to ask ourselves from now on. But I think the timeframe is the question here. It’s definitely not up to scratch right now where it can just take over everyone’s job. But it’s moving so fast that I think what we are facing currently is this wonderful opportunity. There’s a window here for people who decide that they’ve got to push through the fear part, because it’s like standing in front of a tsunami wave. It’s coming, and it’s well and truly on its way, and you can stand on the beach and complain about the fact that it’s going to wipe the whole front of the beach away, or you can figure out a way of riding that damn thing and utilizing it to do the things that you want as a human being going forward and to find a way through this change.

Belinda: So, yes, it’s frightening. Yes, I went through all of those emotions. But it’s also important to acknowledge them and go, “Okay, these are changing times.” Like we’ve all been through COVID, woven through stuff like this is definitely a time of change, and do whatever you need to do to work through that. But I think the only way through is to find the positive angle on this, because once you put yourself into a framework of thinking positively about it and turning off all of those negative thoughts about it, then you see the opportunity. But if you’re always thinking about it in a way like, “Oh my goodness, this is the end of everything,” then it’s impossible to see the opportunity that comes up, because we don’t even know what that opportunity looks like yet. We don’t know what those opportunities are yet, because it’s changing so rapidly. But there’s definitely opportunity there.

Belinda: And where the advantage will be is for everybody who’s a professional in their own right, everyone who does the thing that they do really well. We can utilize this as a tool and offer the best we possibly can to the clients that we have and innovate the ways that we utilize this tool going forward.

Belinda and Anthony discussed how they have used Chat GPT and other AI in their songwriting process. They find it to be a useful tool that can streamline the writing process and inspire creativity. However, they note that the tool is still lacking in its ability to capture the intensity and emotion of lyrics when they are sung.

Belinda believes that the music industry is changing, and the future will be more focused on in-the-moment creation and performance. She predicts that people will gather around artists who can interpret emotions on the spot.

The conversation then turns to clients who believe that the tool can replace human creativity and oversight. Belinda and Anthony suggest making a case for including humans in the process, as the tool is still limited and lacks the full range of human emotions and experiences.

These are definitely important questions to consider when using tools like ChatGPT. Privacy is a major concern for many people, and it’s important to be aware of how data is being used and who has access to it. In terms of bias, it’s true that AI has had issues in the past with reflecting biases that exist in society. One way to avoid this is to be mindful of the data sets that are being used to train the AI models. If the data sets are biased, then the models will be biased as well. It’s important to use diverse data sets and to constantly monitor and adjust the models to ensure that they are not perpetuating biases.

39:43 In terms of privacy, it’s important to be transparent with customers about what data is being collected and how it is being used. It’s also important to ensure that data is being stored securely and that only authorized personnel have access to it. This is something that businesses should take very seriously, as a data breach can have serious consequences for both the business and its customers.

40:06 Ultimately, it’s up to businesses and individuals to be responsible users of technology like ChatGPT. This means being mindful of potential issues like privacy and bias and taking steps to address them. It also means being transparent and honest with customers about how these technologies are being used, and using them in ways that align with the values and goals of the business.

Speaker 1: Utilizing ChatGPT for small business conversations can be difficult due to the biased information it contains up until 2021. Although the platform has worked to balance this out, it’s important to be aware of this bias and ensure that the information output is only as good as the input.

Bek: What about the bias towards white, straight, able-bodied males found in this kind of technology? (NOTE: ChatGPT refused to reframe this question several times)

Speaker 2: There are efforts to counteract bias in the system, but the issue is that we are working in a black box system where we cannot see how inputs are processed. Consumers can create their own versions of ChatGPT to be more aware of the biases and counteract them. This can be achieved through techniques such as weighting data to give more representation to underrepresented groups. However, this is a corporate choice, and the creation of our own versions is currently the main way to counteract bias.

Belinda: There are other options for AI and chat-style AI besides ChatGPT, such as Llama, which can be downloaded on individual computers. This is concerning as people can feed it whatever data set they like, and it’s almost impossible to police. Awareness is crucial, and being on the front end of the wave and developing these technologies for ourselves is necessary to combat the potential negative effects.

Anthony: We are the first generation feeding the machine, and the questions we ask contribute to the data it scrapes. The information we provide now may inform decisions made in 300 years, making it crucial to be mindful of the questions we ask and the data we provide.

Bek: So, you’re saying that you can run the rebellion from the inside?

Belinda: Yes, that’s correct. The questions we asked during our discussion provided valuable information for the AI to make decisions on where to go next.

Anthony: I also think it’s worth noting how much concern there has been about the way the AI has been using data scraped from the internet. In the past, we were all freely giving away this information without much thought to it, but now that AI has the ability to analyze it all in real-time, it’s brought a huge awareness to how powerful this information can be when utilized by AI. The amount of time-stamped data and metadata behind everyone’s actions and purchases creates an extremely powerful way of analyzing human responses and patterns.

Bek: Where can we find out more about you?

Belinda: If anyone wants to find out more about what we’re doing or get a hold of our book, where can they find it? Our book is available on Amazon and Audible, and if anyone needs to contact us, Anthony’s website, anthonysnape.com, has a great contact page.

Bek: As for the chat questions, it seems like some people are excited about the possibilities of AI while others are more sceptical. One person suggested holding on for Google to see how they do it, while another suggested comparing our work with ChatGPT.

Questions from the audience

“I’m going to ask a question from Claire. Before we get Anthony to do his musical thing, Claire mentioned something that made me think about my age and how I learned to type on a manual typewriter, record radio and audio tapes, use phone boxes, and learn WordPerfect. My life is flashing before my eyes, and I’m both excited about the potential of Chat GPT and exhausted at the prospect of teaching it to my mom. Should I ask Chat GPT how to deal with this existential dread and excitement?

Belinda: Absolutely, that’s a great question. You can explain who you are and what you do to Chat GPT and ask for ideas on how to move through this new time of AI. You can ask it to explain things step by step, using terminology that someone from your mom’s generation would appreciate and feel comfortable with. You can draw information from Chat GPT in conversational ways. Push into the machine and discover ways to do that.

Belinda: I want to share with you something that got me really excited and made me realize the world has changed. I asked Chat GPT to write new 10 commandments for living a fulfilled, majestic, purposeful, and love-filled life. Here’s what it came up with:

love yourself and others unconditionally,
follow your heart and listen to your inner voice,
be grateful for the present moment and all that you have,
take care of your mind, body, and soul,
be kind and compassionate to everyone you meet,
be true to yourself and stay true to your values,
take risks and embrace new possibilities,
learn from your mistakes and use them as stepping stones for growth,
be a positive force in the world and make a difference in the life of others,
and live in the present moment and enjoy the journey of life.

I thought that was beautiful and brilliant. Chat GPT can be your best man if you’re struggling with a greeting card or a speech for a wedding.

So, if you’re ever struggling with this kind of thing, give Chat GPT a try. It has all the text from every religion and all the texts available to it.

Anthony: Look, I think that the main thing is, what I really want to make sure that we leave you with is, this is something not to be frightened of, and I think dive in and find something positive out of it, utilize it in whatever way that you need, and be inspired by the potential of what can happen utilizing this leaving that space, because we weren’t fit. Yeah, acknowledge, acknowledge. So there’s how scary it is. But then push on through and find positive ways that we’re using it. Because they’re not going anywhere. It’s not going anywhere, and then help turn your friends around, that are all talking negatively about it and say, Look, I hear you, I feel it. Let’s put that aside. And let’s see what we can do with it. Because it’s positive, it’s here to say, let’s not get washed away in the tsunami of AI. Let’s see what we can do with it.