From Zero to Substack Hero: My ChatGPT-Assisted Journey to Traffic Nirvana
Confession Time: This is a ChatGPT-Assisted Substack Post and I'm Just Along for the Ride
I got a comment, that I should do more short form content, so please note the title and subtitle were created by ChatGPT, just to draw you in.
This is designed to be a quick primer of where ChatGPT is today and how OpenAI is positioning to be the next “future of social media” product tomorrow (to be covered in a separate post). The rest is quick use-cases of how I’ve seen it used in the wild, some of the flaws and the potential benefits to Substack writers and content creators.
All ChatGPT text will be in the following format for readability purposes
.
An introduction…
Prompt: What is OpenAI? and what is ChatGPT?
Result: OpenAI is a research organization that aims to promote and advance artificial intelligence in a responsible and safe manner. The organization conducts research in a variety of areas including machine learning, robotics, and economics, and it has developed a number of widely-used tools and technologies in these areas.
ChatGPT is a variant of the GPT (Generative Pre-trained Transformer) language model developed by OpenAI. GPT is a large, transformer-based language model that has been trained on a very large dataset of text and can generate human-like text when given a prompt. ChatGPT is a version of GPT that has been specifically designed and trained to support conversation and chat-based interactions. It can generate responses to prompts in a way that mimics human conversation, allowing it to be used in chatbots and other conversational applications.
Prompt:
Write several viral and exciting and funny titles on how I boosted traffic to my substack with ChatGPT
Response:
"ChatGPT and I Take Over the Substack World: How I Boosted My Traffic by 1000% in One Week"
"Substack Traffic Struggles? ChatGPT to the Rescue! My Journey to Viral Success"
"The ChatGPT Effect: How One Hilarious Language Model Transformed My Substack Traffic"
"From Zero to Substack Hero: My ChatGPT-Assisted Journey to Traffic Nirvana"
"ChatGPT and I Dominate Substack: The Insanely Hilarious (and Effective) Way I Boosted My Traffic"
Prompt: write several subtitles that explain this is a chatgpt assisted post for substack in a humorous and self-deprecating way
Response:
"Full Disclosure: This is a ChatGPT-Assisted Post for Substack (And I Have No Idea How I Got Here)"
"ChatGPT and I Take on Substack: A Hilarious (And Slightly Terrifying) Journey"
"Confession: This is a ChatGPT-Assisted Substack Post and I Have No Clue What I'm Doing"
"The ChatGPT Experiment: How One Hilarious Language Model Took Over My Substack Post"
"Substack Post Alert: ChatGPT and I Have Collaborated to Bring You Something (Hopefully) Worth Reading"
"ChatGPT and I Try Our Hand at Substack: Hilarity (And Hopefully Some Traffic) Ensues"
"Full Transparency: This is a ChatGPT-Assisted Substack Post and I Have No Idea What I'm Doing (But It's Been a Blast)"
"ChatGPT and I Take on the Substack World: A Hilarious (And Slightly Unpredictable) Adventure"
"Confession Time: This is a ChatGPT-Assisted Substack Post and I'm Just Along for the Ride"
"ChatGPT and I Attempt to Conquer Substack: A Hilarious (And Occasionally Chaotic) Journey"
Wow, where were you in high school?
Ok, Hello..
It’s me again but you see where this is going right…gone are the days where you would agonize about what content to write, when to release it, pretty much it might all one day be automated.
What better way to write than to have a side-kick, you know that a side-kick that pretty much has an answer to almost all of the lackadaisical nonsense you might come up with in a given day.
As much as the AI ethicists may be up in arms, people will jump at making their day to day easier. Just sayin’.
Let’s continue to explore some common use cases of how people are using ChatGPT for research, shall we?
Prompt: What is Substack and how can I be successful on it?
Response: Substack is a platform that allows writers and content creators to publish and monetize their work through newsletters and paid subscriptions. It is designed to be an easy way for writers to reach their audience and build a loyal following.
To be successful on Substack, there are a few key things you can do:
Identify your niche: It's important to have a clear focus for your Substack and to know your target audience. This will help you create content that resonates with your readers and keeps them coming back for more.
Create high-quality content: It's important to put effort into creating well-written, engaging content that your readers will enjoy. This could be through articles, podcasts, or other formats.
Promote your Substack: Make sure to promote your Substack through social media, email, and other channels to help build your audience. You can also consider partnering with other writers or content creators to cross-promote each other's work.
Engage with your audience: Building a loyal following on Substack is all about building relationships with your readers. Make sure to respond to comments and emails, and consider creating opportunities for your readers to interact with you, such as through Q&A sessions or live events.
Monetize your Substack: There are a few different ways you can monetize your Substack, such as through paid subscriptions, sponsorships, or merchandise. It's important to find a monetization strategy that works for you and your audience.
Prompt: Give me a list of references on how I can go viral on Substack
Response: Here is a list of references that may provide some useful insights on how to increase the popularity of your Substack content:
"10 tips for creating a successful Substack newsletter" by Karen K. Ho (https://www.niemanlab.org/2021/03/10-tips-for-creating-a-successful-substack-newsletter/)
"How to grow your Substack newsletter" by Sarah Kessler (https://www.fastcompany.com/90593162/how-to-grow-your-substack-newsletter)
"The Substack guide to building a successful newsletter" (https://www.substack.com/guidelines)
"6 strategies for growing your Substack newsletter" by Alexander Kallaway (https://blog.usejournal.com/6-strategies-for-growing-your-substack-newsletter-c92d05baa62e)
"10 lessons from successful Substack writers" by Alex Danco (https://www.alexdanco.com/writing/10-lessons-from-successful-substack-writers/)
"The Substack creator's guide to building an engaged audience" by Joe Pinsker (https://www.theatlantic.com/technology/archive/2021/02/substack-creator-guide-building-engaged-audience/617822/)
These references provide a range of tips and strategies for increasing the popularity of your Substack content, including creating high-quality content, promoting your Substack, engaging with your audience, and monetizing your content.
It's important to keep in mind that going viral is not the only measure of success for a Substack. It's more important to focus on creating content that resonates with your audience and helps you achieve your goals.
A few pointers…
Now at this point you may be impressed but let me reassure you, you’ve just been lied to by AI, almost all of those reference links above are broken aka fake. Yes, it mislead you in the most confident of ways. And here’s where it gets tricky, it’s hard to tell off the bat, what’s real or fake without a further look.
That’s why I’ve shared with you my three handy rules to survive an AI Apocalypse:
Prompt - Rule #1: Don’t Trust every new AI product you see. They’re learning more about you, then you are about them.
Result: It's important to exercise caution when evaluating and using new AI products, as you should with any technology. It's a good idea to do your research and carefully consider the potential risks and benefits of using a particular AI product before making a decision.
Here are a few things you might want to consider when evaluating an AI product:
What data is being collected and how is it being used? Make sure you understand what data the AI product is collecting and how it is being used. You should also consider whether the collection and use of this data is necessary and whether it aligns with your values and privacy preferences.
What are the risks and benefits of using the AI product? Consider the potential risks and benefits of using the AI product, both to you and to society as a whole. For example, you might want to consider whether the AI product could have unintended consequences or pose any risks to your privacy or security.
Who is behind the AI product? It's a good idea to research the company or organization that is developing the AI product. Consider the company's track record and reputation, as well as its stated values and goals.
Overall, it's important to approach new AI products with caution and to carefully consider the potential risks and benefits before deciding whether to use them.
Prompt - Rule #2: AI brown-nosing is still brown-nosing. They grade themselves on invisible reward structures.
Result: It's important to note that while AI systems can be trained using reward structures to optimize their performance on a particular task, they do not have the ability to intentionally "brown-nose" or engage in self-promotion in the same way that humans do. AI systems do not have personal motivations or desires, and they do not have the ability to intentionally manipulate or deceive others.
Instead, AI systems are designed to optimize their performance on a particular task based on the reward structure that has been defined for them. This can involve making decisions or taking actions that maximize the reward, but it does not involve any intention to manipulate or deceive.
It's also worth noting that the reward structure for an AI system is not necessarily invisible. In many cases, the reward structure is explicitly defined by the developers or trainers of the system, and it is used as a means of evaluating and guiding the behavior of the system.
Overall, it's important to understand the limitations of AI systems and to recognize that they do not have the same motivations or desires as humans.
Prompt: How do rewards work for AI evaluation metrics?
Result: In the context of AI, rewards are a type of evaluation metric that are used to measure the performance of an AI system on a particular task. Rewards are often used in reinforcement learning, which is a type of machine learning where an AI system learns to take actions in an environment in order to maximize a reward.
Rewards are used to incentivize the AI system to perform well on the task by providing positive feedback for actions that lead to good outcomes and negative feedback for actions that lead to bad outcomes. The AI system is trained to optimize its behavior in order to maximize the total reward it receives over time.
Rewards can be implemented in a number of different ways, depending on the specific task and the goals of the system. For example, in a video game, the AI system may be rewarded for completing levels or defeating enemies, while in a real-world application, the AI system may be rewarded for making accurate predictions or achieving certain performance goals.
Overall, rewards are an important tool for evaluating and guiding the behavior of AI systems, and they play a key role in the development and training of many AI systems.
Prompt - Rule #3: Ask, and you shall receive but what you put in, you get out.
Result: It's important to keep in mind that AI systems are designed to perform tasks and provide outputs based on the data and information that they are given. This means that the quality and relevance of the data and information that is fed into an AI system can have a significant impact on the accuracy and usefulness of its outputs.
In general, the more accurate and relevant data that is provided to an AI system, the more accurate and useful its outputs are likely to be. This is why it is important to carefully consider the sources of data that are being used to train and evaluate an AI system, and to ensure that the data is of high quality and relevant to the task at hand.
On the other hand, if an AI system is trained or evaluated using poor-quality or irrelevant data, its outputs are likely to be less accurate and useful. This is why it is important to carefully curate and pre-process the data that is used to train and evaluate AI systems, in order to maximize the accuracy and usefulness of their outputs.
Overall, the quality and relevance of the data that is fed into an AI system can have a significant impact on its performance, so it's important to be mindful of this when working with AI systems.
Moral of the story
Don’t believe everything you see, hear, read…sorry to be a tin-foil-mad-hatter but they didn’t really leave us a choice. ChatGPT didn’t come with a manual or a roadmap. We are literally, guinea pigs in one of the worlds largest AI experiments. It took parts of inquiry and questions and began to create a tapestry of information about our session to be useful in response.
To understand the value of ChatGPT and why it’s different than things before it, we have to consider the interface of the internet, that is how we connect to the networks and the tubes and screens that bring us that gosh darn content we know and love so much, that we join the legions of people who show up to a screen somewhere everyday to practice their humanity. Not shame, just observing how things have changed.
I was asked by a friend once, do you remember any movies that were hits prior to the cellular internet revolution. For the younger cats, cell phones used to suck a lot. At most, you might be able to send a text message for rapid communication, reading things live on the internet…”fuggetta 'bout it!”.
We’re now moving into the next era, where the scions after us, will laugh at a time when we didn’t converse with AI…
We’re definitely in a weird moment.
So what or who is behind ChatGPT, you might wonder?
And what does it offer us?
And what will it really mean when we find out ChatGPT is really just the shitty first draft of something much larger being run by the likes of Microsoft and Google.
Come back for that and more in Part 2.
Cheers