They expect full automation right out of the gate – stellar copywriting or content writing, without fine-tuning the model, providing sample copy, instructions, or anything.
So of course it won’t get great results…
Currently, ChatGPT is best for augmentation, not full-scale automation.
Yet when it doesn’t meet unrealistic expectations…and despite the fact that AI has appeared out of nowhere to and can suddenly generate infinite amounts of text that is better than many humans…the attitude has shifted from one of amazement to “oh, it sucks now so it will always suck.”
But AI hasn’t even gotten started yet.
This is literally just the first inning…and, importantly, it can already fully automate many tasks better than humans, at a much lower cost, and much, much, much faster.
If you are not prompting it correctly…and if you’re using ChatGPT instead of the OpenAI API or the Playground Assistants…and if you’re not feeding it sample copy, then of course what comes out is going to be weird.
(I actually think OpenAI did that on purpose, but that’s a different topic.)
But you don’t just take the words it gives you and then stop.
You can refine the output repeatedly, using things like samples. And:
Tell it what readability level you want to use (e.g., 8th grade reading level)
Tell it what phraseology to avoid and what phrases to use
Tell it to be more concise, more formal, less formal, use third-person, etc.
You can give it a lot of different instructions.
And you can provide it with copy to rip off…er, “sample.”
When you start doing things like this, the copy gets a lot better.
The unrealistic, unfounded expectation is that ChatGPT should be able to magically create better copy than humans, while leaving humans completely out of the loop.
But with a lot of copywriting, it’s about a spectrum of augmentation to full automation.
You just follow the trend line:
I used GPT-3, and then I used 3.5, and then I used 4.
And it keeps getting better and better and automating more and more.
And Sam Altman recently said that the next model is going to be way better.
So, since ChatGPT and Midjourney and Runway ML literally came out a year ago, I think we should stop acting like their automation impact has leveled off.
We don’t know that.
I think most people should listen to the only people who are experts: the people developing actually developing AI.
And these people, like Sam Altman and Kai Fu Lee, literally say that we’re all going to be out of work.
If they’re right, there is a possibility that today’s white-collar jobs will go the way of the farmer.
99% of them will disappear and be replaced by machines.
And what will be left?
“Gold-collar workers”?
I don’t know, but I’d rather be safe than sorry.
P.S.
An example of how to use AI to augment your writing (rather than attempt “full automation” without proper preparation):
I spoke this post into a Google Doc, which produced a big long incomprehensible text stream
I gave that long text stream to a Custom GPT that I built specifically for cleaning up transcriptions
Once it was tidied up, I copyedited the transcript
Then I gave the transcript to another Custom GPT I made specifically for generating Midjourney images
The result:
A complete direct response copy-style post in a fraction the time it would’ve taken me to write it manually
A lighter cognitive load (very important for writers)
40 Midjourney images to choose from (no need to search for or pay for stock photos)
Along with other white collar job roles, I expect AI to drastically change the way writers work, their role in the company, and the value they add.
While many may not like it, there’s not much we can do about it.
The only solution is to adapt…and because AI is moving at such a breakneck pace, this means adapting as soon as possible.
Atlassian, a leading software company known for products like Jira, Confluence, and Trello, has recently made significant changes to its workforce and its software suite, both of which mirror wider trends in the technology sector. These changes include substantial layoffs, a shift from server-based to cloud-based solutions, and a growing emphasis on artificial intelligence (AI) and automation.
In 2023, Atlassian announced the layoff of approximately 500 employees, which constituted about 5% of its total workforce.
The decision, which they described as a rebalancing effort, was not primarily financially motivated but aimed at refocusing on areas with growth potential. The company’s leadership emphasized that this was a strategic move to prioritize cloud migrations, IT service management, and serving enterprise customers through cloud solutions.
Atlassian is ending support for the Server editions of its software, a move that highlights the industry-wide shift away from data centers and on-premises offerings.
The company plans to retire its self-managed server offerings in February 2024, a decision they announced in late 2020. This transition marks a significant change in their business model, and they are encouraging customers to migrate to cloud-based or Data Center products. Atlassian is offering tools and resources to facilitate this migration, such as enhanced security, scalability, automatic updates, and access to the latest features and functionalities.
More recently, Atlassian also announced that they are integrating AI assistants into their software.
The AI features are set to offer substantial benefits, such as improved productivity and enhanced features for their customers. The integration of AI will be used to:
Generate insights
Automate routine tasks
Create more intuitive user experiences
As mentioned, these moves by Atlassian reflect broader trends in the technology industry.
The shift from data centers to cloud computing is driven by the many advantages of the cloud, such as
Scalability
Flexibility
Cost-effectiveness
Improved security
And the rise of AI and automation signifies a shift towards more intelligent and efficient business solutions.
The Accelerating Automation Trend: Doing More with Less
Atlassian’s recent layoffs, the transition from server to cloud, and the focus on AI are indicative of the larger shifts occurring in the technology industry.
While Atlassian’s move seems to be proactive rather than reactive, they are certainly not the only ones restructuring in response to tech-driven innovation:
Intel continues cutting jobs as competitors, such as Nvidia and AMD, eat up the market for AI and, soon, desktop CPUs
In the coming months and years, we should expect to see more layoffs, more automation, and more disruption.
Given the speed at which AI is disrupting the workplace and the chip market, I personally think companies should immediately rethink the way they operate, hire Chief AI Officers (CAOs or CAIOs), build innovation hubs, and start adopting AI tools immediately.
So, what is the future of content marketing now that ChatGPT and generative AI have hit the scene?
For many content marketers and marketing creatives, GenAI looks like it might turn the entire world upside down:
ChatGPT can write in any language, including programming languages
Midjourney can create stunning images of any subject in any style
Runway ML and Pika can turn images or text prompts into video clips
Google and Meta are using AI to create and optimize ads
These are a few of the many examples of how AI is invading the workplace.
And while some may say we can never know what the future holds, I completely disagree.
I think the writing’s on the wall.
In this article:
I’m going to walk through four trends that content marketers can expect in the coming months and years, some of which are well underway
I’ll show what a weird world we’re moving into
And I’ll explain what content marketers, content creators, and marketing creatives need to do in order to stay relevant and competitive
To start, let’s look at the first trend:
Trend 1: Content Will Become a Cheap Commodity
Less than a year since ChatGPT hit the big leagues, it was already able to generate top-quality writing in seconds.
Anyone can sit down with ChatGPT+, feed in a few keywords, and churn out mass content that is better than 80% of human SEO content writers (my ballpark estimate).
Some SEO professionals suggest that, to combat this trend, Google is already giving more weight to backlinks.
Unfortunately, this means that search results are going to be skewed more in favor of “authoritative” websites instead of relevant content.
And, like any newspaper or advertising-driven publisher, this means that Google will favor bigger companies with bigger ad accounts.
Solopreneurs, bloggers, and small businesses with smaller SEO budgets will have a harder time ranking because they need to purchase backlinks, guest posts, articles, and advertising.
But SEO content is only one of many types of writing ChatGPT excels at.
I’ve used it to write:
Short stories
Novel outlines
Poetry
Songs
Curriculums and lesson plans
Metaphors, parables, and analogies
In short, anything you can think of, it can write…and since text is also the language used to feed image, video, and code generation, we’re looking at a very daunting pile of job tasks that stand to be hit by the GenAI automation wave.
Trend 2: Web Content Such as Blog Articles Will Disappear
On top of the challenges of SEO, attention spans have been shrinking for a while now.
“No one reads anymore,” complain many readers and writers.
And it’s true.
With the rise of video, social media, Kindle, podcasts, and other new media formats:
Attention spans have withered
People read less
They spend more time in digital worlds than the real world
We are leading busier lives
There’s more and more noise competing for our attention
It should be no surprise, therefore, that text is being competed away by newer, more stimulating forms of content.
And when people do read, they don’t have the patience or the time for fluff, creative wordplay, or writing that taxes the brain: both people and Google want well-organized content that is clear, skimmable, scannable, and easily summarizable.
With the rise of video and the Metaverse, I fully expect “boring old text” to get drowned out by the noise of the AI-generated hyperreal.
Content marketers and agencies should take notice here, because this will have a significant impact on business, revenue, and marketing models.
Trend 3: AI Controls Your Content
The third trend that interests me is the fact that AI has become the new gatekeeper of all content.
Now, Google is using AI to generate answers to search queries underneath the Google search box on the search results page.
Known as the Search Generative Experience (SGE), Google’s new AI search feature summarizes web content into answers that appear above search results.
Useful for informational queries, SGE may provide a better user experience in some cases, but it also has a few unfortunate side effects for marketers, businesses, and anyone looking to get noticed through Google:
SGE pushes search results further down the page and, as a result, content becomes further obscured
Google now has even more control over what information is presented to users and the way it is presented
Rather than becoming a “search engine,” Google’s function has now changed to that of an “answer engine”
For marketers, one key takeaway is that SGE, virtual assistants, and other AI-powered search engines will be the new filters for marketing content.
And since AI firmly remains under the thumb of its creators, whether that is Google, OpenAI, Amazon, or Apple, these companies will determine whether or what content reaches the global audience.
Marketers should therefore treat these new AI-driven search engines as a platform, like Facebook, which has a particular audience, internal biases, and rules.
So if you want to reach your audience, you’ll need to understand and abide by these rules.
Trend 4: Say Goodbye to Generic Broadcasts, Say Hello to Auto-Personalization
There is another side effect of GenAI that many have yet to grasp: the ability to use GenAI to programmatically generate more GenAI.
For instance, with the right data, you can automatically generate text, which can be used for image generation, video generation, and 3D environment generation across virtually any channel, including:
Advertising
Email
Websites
Social media
Apps
The metaverse
Imagine a world where:
Emails are 100% uniquely written for you and no one else
Website design, imagery, copy, and even the UI is tailored to you
Video ads are crafted specifically to convert you
Ads that are dynamically generated based on your recent purchase history
ChatGPT and other virtual assistants automatically adjust their words, tone, and approach based on your mood
Yes, it all sounds futuristic, yet the technology for this level of automated hyper-personalization already exists.
When you look at all of the trends covered so far, we seem to be witnessing a convergence of disruptive trends that could cause some serious upheaval in the next few years.
When GenAI comes into full force, what will the world of content marketing look like?
Content Will Be Dynamically Generated, 100% Uniquely Personalized, and Mixed Reality
[Don’t read the next section unless you want to read spoilers about the movie Solaris]
Nope, not from the movie. This, like all images in this post, are from Midjourney.
Solaris is a planet that can read your thoughts and turn them into physical reality.
Featured in a movie of the same name, the movie follows a psychologist, Kris Kelvin, sent to a space station orbiting a mysterious planet.
There, he encounters manifestations of his deepest memories and guilt, such as those of his dead wife.
The story unfolds as Kelvin grapples with reality, grief, and the strange planet, which, as we gradually discover, seems to read and materialize the crew’s thoughts.
It creates physical, real-life replicas of the crew’s deepest memories and regrets.
Their interactions with these manifestations lead to psychological and emotional turmoil, challenging their perceptions of reality and sanity.
In the concluding scene of “Solaris,” Kelvin returns to his family home and meets his father.
But it’s apparent that something is off.
There’s rain inside the house and his father is behaving oddly, which suggests that Kelvin isn’t home at all…he’s still on the constructed reality of Solaris, simply dreaming that he’s returned home.
This ambiguous ending leaves us questioning the nature of Kelvin’s reality, a scenario that we all may be facing…and probably a lot sooner than we think.
Can Marketers Crush It With Content in a GenAI World?
Like Solaris, AI will be able to read our thoughts and emotions, then use those to dynamically generate hyper-personalized content that’s tailored specifically for our psychology.
Don’t believe me?
Why hire models and stock photographers anymore when Midjourney v6 can do this?
Today, it can already perform sentiment analysis on text, understand the feelings in our voices, and recognize lies in facial expressions.
And it can generate text, images, and video that’s indistinguishable from reality: a.k.a., deepfakes.
What happens when it can auto-generate this type of content dynamically?
Sure, it could plunge us all into a Solaris-style world where we have trouble distinguishing dreams from reality.
But existential Matrix-style philosophizing aside, let’s focus on the practical impacts this will have on content marketers and the marketing industry:
AI will enable extraordinary levels of creative expression in new and old media alike
People will gravitate towards new AI-created media, like video, AR, VR, virtual assistants, and games
Yet with GenAI, the production of text, images, video, and 3D environment development is being commoditized
Content marketers will need to reskill, adapt to new workflows with AI tools, and find new ways to add value that can’t be replicated by AI
Many experts believe that as a result of AI automation, competition will skyrocket and only the hardest-working, most skilled, and most talented will survive
I, for one, see the burgeoning GenAI megatrend as both an opportunity and a crisis. It’s an opportunity for those who adapt quickly, seize the moment, and jump headlong into the AI race. But it’s also a crisis, because AI will almost certainly automate more jobs more quickly than any other trend in history.
But I don’t want to mire myself in the automation discussion.
I’d rather focus on what marketers can and should do to succeed:
Constantly use and experiment with AI tools
Use multiple AI agents to streamline and amplify your content production pipeline
Reskill, upskill, cross-skill, de-specialize, and become a T-shaped professional
Hit where the puck is going to be, not where it is now
Because the world is moving through a disruptive transition, it can be difficult to know what moves to make or when to make them.
But since AI is the very force driving the next wave of tech disruption, there’s only one conclusion:
To Thrive, Content Marketers Must Be Aggressive Adopters of AI
Content marketing is at a crossroads.
AI’s ability to generate human-grade content poses a significant challenge to marketers and creatives, but it also opens doors to unparalleled opportunities.
I could wax philosophic with more generic ChatGPT-like fluff and say things like, “To not just survive but thrive in this new era, marketers must embrace the change.”
And that’s true.
But I prefer practical tools and tactics, which is why I’ll be focusing a good deal of energy on the “how to” of AI workflows, using tools like ChatGPT, Custom GPTs, LangChain, Midjourney, Runway ML, Pika, and more.
If you haven’t already, subscribe to my newsletter to stay tuned.
And if you would like support with your own content marketing efforts, email me at nathan@nathantwarne.com.
The right LLM-based app, I believe, can give content production a massive boost, in terms of both quality and quantity – while simultaneously saving time and money.
If you’re an SEO, content marketer, or agency, then a well-designed LLM-powered app can allow you to “write once and deploy content anywhere.”
In the coming months and years, I think there’s an 80%+ chance that the entire digital landscape will be cannibalized by AI: LLMs will underpin all software and everything will be connected by APIs.
Of course, for now, what we’re looking for is practical solutions.
Cutting Content Marketing Time in Half with ChatGPT
Now, with ChatGPT, you simply need to write (or even speak) your content, then it can:
Copyedit and proofread
Perform more research
Make suggestions
Draft up and edit copy for:
Blogs
Social media
Newsletters
Marketing copy
Sales copy
Books
Provide the output on screen or as a downloadable file
With Custom GPTs, you can provide custom instructions and basic functionality, including web browsing, coding, and image generation.
I’ve created a few that I use for:
SEO and data analysis
Web research
Drafting up content
Generating Midjourney prompts
This, however, is just the tip of the iceberg of what’s possible.
Building an LLM-Powered Content Marketing App with LangChain
With frameworks like LangChain, you can create an app that:
Pulls in data from a wide variety of sources, including Google (via APIs like SERP API), email, web scraping, and SEO tools
Stores that raw data and analyzes it
Searches for more relevant data
Summarizes the content
Extracts topics and keywords
Provides SEO content recommendations
Drafts up content for channels of your choice
Create SEO-related content, such as meta descriptions and image alt text
Recommend and insert SEO-related anchor text
Use APIs to schedule posts across platforms
All of this is possible right now with services such as the OpenAI API and frameworks such as LangChain, so it’s only a matter of time before we see content marketing and SEO tools hit the market…and massively disrupt the way we work.
Until then, however – and partly because of the impending threat of disruption – I think it’s critical that we all innovate as aggressively as possible with these tools.
AI agents interact with LLMs to autonomously or semi-autonomously perform tasks and achieve specific goals. ChatGPT’s recently released “Custom GPTs” are an early example of AI agents.
What Are AI Agents?
In a nutshell, AI agents:
Take a set of custom instructions, such as a custom templated prompt
Have access to a set of tools, such as web browsing, an API, or data analysis capabilities
Can act autonomously or semi-autonomously to perform specific tasks
AI agents can perform tasks such as:
Web research
Data analysis
Writing and editing
Sending emails
Scheduling meetings
Performing tasks inside software applications via APIs
Pulling data from APIs
In my opinion, agent apps such as ChatGPT’s Custom GPTs will become the norm in the next few years.
Today, Custom GPTs can take custom instructions and use a limited set of features, such as web browsing, data analysis, and image generation.
But in the near future, we can expect much, much more.
Systems of AI Agents
Once AI agents become integrated with APIs, they will be able to connect to apps directly and perform tasks inside that app.
Imagine being able to give commands or ask questions like these:
“Send a text message to Bill and tell him I’m going to be late to the BBQ.”
“Email Alice and ask her if she’s available to meet at 1 pm tomorrow.”
“Create a new ClickUp task in XYZ space with XYZ description, estimated at XYZ hours, due in 3 days.”
“I’ve emailed you a spreadsheet. Analyze it and tell me what it’s about.”
These are just a few very minor examples of what AI agents will be doing for us in the very near future.
Their real power comes when you put multiple agents together into one single system.
For instance, let’s say you want a set of AI agents that can help with SEO.
Agents could specialize in tasks that include:
Domain research
Keyword research
Backlink research
Content research
Competitor research
Data analysis
Content creation
The more agents you have, the more tasks they can perform, and the more of your job you can automate.
AI Agents and the Future of Work
For early adopters of AI, this is great news, because it means you can 10X your productivity with the right system of AI agents.
When ChatGPT first got big, I spent weeks locked up in my room exploring its capabilities, researching LangChain, and trying to figure out where this technology is headed.
To me, it seems clear that systems of AI agents are the “next big thing” in AI, so this is where my efforts will be focused.
Despite the hype surrounding AI, a writer for The Guardian says we still don’t know what the long-term impact of AI will be.
True.
We don’t know what the world will look like in 20-50 years.
But the experts…those building AI, like Sam Altman and Kai Fu Lee…say AI will automate practically everything and that financial solutions like universal basic income (UBI) are the answer.
But even today, the last day of 2023, barely a year after ChatGPT was introduced, it’s clear that AI is already fueling massive automation.
Google, for instance, recently announced that, thanks to AI, a number of jobs in their 30,000-person ad sales department are now redundant.
Who knows how many jobs will be on the chopping block? Hundreds? Thousands?
I’ve always been a believer in automation companies like UiPath, AutomationAnywhere, and WalkMe, but now I’m even more bullish.
If any company stands to profit from an automation revolution, it will, of course, be automation companies.
The performance of companies like UiPath and WalkMe hasn’t been so great the past two years, but they’ve been skirting what I hope is the bottom of a couple-year dip.
They have already been pioneers of innovation, developing functional, useful chatbots well before OpenAI came along. GenAI might be the catalyst that enables them to create new products that launch their companies into the stratosphere.
ChatGPT and LLM-based apps can dramatically boost productivity, efficiency, and output.
With the right prompts, tactics, techniques, and strategies, you can do more, work faster, and get ahead of your competitors…and your peers.
If you like the idea of automating your job with ChatGPT but aren’t sure just how to use these tools, keep reading.
This guide offers 20 essential tips to enhance your productivity and interaction with AI.
1. Understand the Context Window
ChatGPT can only remember so much.
And the longer you talk to it, the more it forgets.
Understanding the context window is crucial when working with models like ChatGPT, which have a limited “memory” for text:
Provide context to get more relevant responses
Break up longer texts into shorter messages
Place instructions at the beginning or the end of a message (I saw one experiment suggesting that LLMs tended to better “remember” instructions at the end of the message)
Periodically ask the AI what it remembers about your conversation – this will give a sense of what “context window” means
Over time, the context window and AI performance will continue to grow, so don’t get discouraged just because it can’t copyedit a 2,000-word article or write a complete research paper today…we’ll get there eventually.
2. Crafting Effective Prompts
Go into great detail with your prompts, then copy & paste them into custom instructions
Crafting effective prompts is the foundation of success with ChatGPT and LLMs.
Learn to phrase prompts clearly and specifically and provide specific examples and the desired structure for your output.
Include things like:
The AI agent’s role
A goal or purpose
Your role
What level of learner you are
The tone/voice/reading level to write in
Examples of the type of output you’re looking for
How you want the output formatted
Refine your prompts and consider using custom GPTs with their instructions for repetitive tasks that need consistent output.
3. Specifying Tone of Voice
Specify the tone of voice to avoid copy that sound robotic or weird. You can even copy and paste large chunks of copy as a sample.
With the right prompt, examples, and coaching, you can align the AI’s output with your expectations. Test different tones, and use prompt examples available online, especially those designed for copywriting and writing.
This will help you craft responses that resonate more effectively with your intended audience.
4. Zero-shot Learning vs. Examples
Understanding the difference between zero-shot learning and providing examples is essential if you want to work with ChatGPT and LLM-based apps.
Zero-shot learning means that the AI will respond without examples to work from – this is frequently what creates so-called “bad” output.
Providing a few examples, however, is known as few-shot learning.
Doing this is kind of like “mini-training,” helping the AI understand your request more accurately.
5. Formatting Instructions
Clearly outline the desired information presentation format – for instance, tell ChatGPT that you want bullet lists, markdown, plaintext, headers, subheaders, and so forth.
You can also use “variables” in your copy.
For instance, you can create a prompt that looks like this:
Extract the product information, benefits, & features from the text above. Give them to me in a bullet list that looks like this:
[Feature]. [Benefits of feature].
[Feature]. [Benefits of feature].
…
By being specific about the format, you ensure that the AI delivers the information in a way that suits your expectations the first time around.
6. Recognizing Built-in Biases and Politics
An LLM app comes with its own set of inherent biases and political leanings programmed into it.
Keep this in mind and adjust your prompts accordingly.
Note that although adding specific instructions for each prompt can be tedious in ChatGPT, custom instructions and Custom GPTs can help streamline this process, since you’re only giving instructions once.
7. Using Multiple Custom GPTs
Custom GPTs are just the beginning of AI Agent “Systems,” which will completely transform the way we work.
One way to automate multiple tasks is to give a lot of instructions to one agent or Custom GPT.
Another is to use multiple Custom GPTs – which is what I do.
Experiment with various GPTs to find the one that suits your needs best.
As of this writing, ChatGPT is a frontrunner in the LLM field, so many…including me…stick with OpenAI.
This means I can focus on my own workflows and tasks, instead of tinkering around with different GPTs – after all, they’ll all achieve parity at some point.
8. Giving Custom GPTs Access to Tools
Enhance AI capabilities by allowing access to various tools.
Custom GPTs can use functions and tools that enable them to perform diverse tasks, such as connecting to an API to check the weather, write code, or generate images.
The more tools an agent has, the more it can do.
Experiment continuously to discover how much you can automate.
The OpenAI Playground’s “assistants” and platforms like LangChain allow even greater creativity and functionality, since you can create multiple agents with a wide range of capabilities.
9. Setting Clear Goals
In your prompt, clearly define your goal and the agent’s role.
Each agent or custom GPT needs a purpose, so make that clear by saying things like:
“You’re a coder with 30 years of experience.”
“Teach to a university-level learner.”
“Your sole job is to proof for errors, correct grammatical mistakes, and fix punctuation. That’s it.”
This ensures the agent actually does what it needs to do.
10. Providing Continuous Feedback
Regularly coach the agent to refine its performance.
Once you find the output you like, create a prompt then and there.
Or ask the AI to generate a prompt that will reproduce the output you’re getting, then put that prompt into the custom instructions.
11. Stay Updated on AI’s Developments
AI is changing rapidly and it won’t stop anytime soon.
Though a lot of people aren’t paying close attention, that doesn’t mean it’s stopped moving.
It’s evolving quickly, and one day we’ll wake up to a vastly different landscape.
Since today’s early adopters will be tomorrow’s winners, it’s crucial to keep up with AI as aggressively as possible.
Follow blogs, YouTube channels, social media accounts, and influencers – and, most importantly, use AI daily.
12. Understand Limitations
AI is not omnipotent.
But it’s also not as useless as some people believe…
Some people outright dismiss ChatGPT as unintelligent after a few unsatisfactory responses, but others are achieving unbelievable results.
SEO professionals, for instance, are among the most creative and innovative tech users in the world, and many of them are 10Xing their productivity with AI.
13. Consider Privacy and Stay Compliant
When working with AI, especially in a business context, understand that the data you submit might be confidential and could be used for AI training.
Don’t submit any information you wouldn’t want companies like OpenAI or Microsoft to access.
Being mindful of data privacy policies and the information you share is crucial for maintaining confidentiality and trust.
14. Experiment Small, then Progressively Overload
The context window for AI has a finite size, so it’s best not to overwhelm it with too much information at once.
Start with smaller tasks and gradually work your way up to understand its capabilities and limitations better.
Experiment with content size and use multiple messages when necessary to find the right balance for effective communication.
15. High-Level Prompt Strategies
Prompts will make or break the output.
Mastering effective communication with LLMs is key, and techniques like Chain-of-Thought prompting, Tree of Thought, and Chain-of-Verification can significantly enhance performance and reduce errors.
Chain-of-Thought:
think step-by-step through the process
Tree of Thought:
Imagine three different experts are answering this question.
All experts will write down 1 step of their thinking,
then share it with the group.
Then all experts will go on to the next step, etc.
If any expert realises they’re wrong at any point then they leave.
The question is…
Chain-of-Verification:
[Initial prompt]
List capitals of countries in Asia.
[Initial response]
Bangkok, Tokyo, Beijing….
[Question to verify understanding/correctness]
[Response]
[Final answer]
Once you understand and apply these techniques, particularly in custom instructions, you can improve your output significantly…and consistently.
16. Experiment with Different Strategies
Continuously innovate with prompts, tools, and models to keep pace with the rapidly changing industry and technology.
As the tech evolves, so should your strategies:
Test prompts
Experiment with your own Custom GPTs
Experiment with those made by other people
Try new tools and software
Make your own tools and software
The more you learn, the better off you’ll be when the next “AI tidal wave” hits.
17. Monitor AI Evolution
Stay informed about advancements in AI and how they might affect or improve your workflow. Subscribe to:
YouTube channels
Email newsletters
Blogs
Social media accounts
Also, attend events, conferences, and seminars.
Focus on AI related to your area of expertise.
18. Develop a Learning Mindset
AI is poised to significantly impact the world, so getting ahead now is crucial.
Invest in learning and become an expert in AI within your industry:
Take courses
Get certifications
Broaden your skill set, especially if AI is threatening your job
Become an expert and get recognized as one
The more you can apply AI in your day-to-day job, the better positioned you’ll be for future opportunities.
19. Collaborate with Others
The AI community is overflowing with creativity and innovation.
Engage with peers, join online communities, and follow influencers who are well-informed about AI.
Mainstream media may not always provide the most accurate picture of AI’s potential – I often see articles on mainstream sites by people who fail to see its value.
So stay on the right side of history: seek out those who truly understand its impact and its future.
20. Ethical Considerations
AI is opening a portal to a very weird world. Let’s make sure we get it right.
Be mindful of the ethical implications of your AI interactions.
As AI technology becomes more pervasive, it’s crucial to consider how it’s used and the potential consequences.
Reflect on:
Privacy
Data security
Compliance
The value you’re bringing to the world
AI is neither good nor evil.
It is an amplifier that can be used for good or for evil, so I think it’s useful to consider how – or whether – you are adding value to the world.
Are you developing an AI influencer that is getting kids hooked on dopamine and a brand that does nothing but drain wallets?
Are you using deepfakes to scam people?
Are you building an empire on teaching get-rich-quick-with-ChatGPT schemes?
Or are you building something constructive?
Conclusion
ChatGPT and other LLM-based applications are set to transform the world, one job task at a time.
We can spend forever debating the value in that transformation, but at the end of the day you’ve only got one choice: adapt or get left behind.
Will Ginkgo Bioworks become “the Amazon” of Synthetic Biology?
The founder, Jason Kelly, certainly thinks so.
Ginkgo Bioworks (NYSE: DNA), has big partnerships with companies like Moderna and Bayer, among many other prominent names.
They’ve also received funding from the likes of the Bill & Melinda Gates Foundation (though it’s just a very modest $2 million or so, a far cry from the vast sums they dumped into Moderna).
Fast time-to-market has always been important in content marketing, but AI has launched us into a completely new marketing world.
In 2022, copywriters could spend hours brainstorming, researching, writing, and editing a piece of copy – and then pass it along to clients, supervisors, or other stakeholders to review and send back for more edits.
Depending on the workflow, the total process could take days…or even weeks.
But AI has changed all that.
Now, in 2024, tools like ChatGPT can cut significant time off the content development timeline, while simultaneously improving quality.
One implication of this is that speed will become even more important than it already is.
When AI can write at the speed of light, you can’t afford to spend ages on a single post, because your competitors will complete ten articles, emails, or ads in the time it takes you to complete one.
Marketers are just now beginning to realize the need for more speed and efficiency:
Perhaps it’s because I’m an early adopter (I’ve been using GenAI copy tools since early 2021), but given the breakneck pace of innovation in AI, I found this a little bit surprising.
2.5 quintillion bytes of data is being created every day
60% of that is uploaded by humans
Between 2010 and 2021, the total amount of data created and replicated globally grew from 2 zettabytes to 79 zettabytes
In this noisy online world, quantity is what helps you get noticed, and quality is what helps you position yourself as a trusted authority.
But those statistics are current as of 2021, so just imagine how much things have changed – and, more importantly, how much more they will change with AI.
Content marketers will be 5-10X as productive and 5-10X fast, while producing higher-quality work at a lower cost.
Results Will Be Harder to Come By – And More Important Than Ever
While the advent of AI has changed the game forever, the bottom line for all marketing remains the same: getting results.
Because every marketer and content creator will be lifted to a new level of quality with AI, we now have a new, higher baseline to operate from.
In my opinion, low-quality SEO content, spammy-sounding sales messages, shortcuts, and haphazard marketing campaigns just won’t cut it anymore.
Some may say this sounds a bit melodramatic.
But this trend towards increased quality, quantity, and competition has been ongoing for years.
The Information reported that Google plans to restructure its 30,000-person ad sales unit.
Per Ars Technica, however, it’s a “restructuring” that will “consolidate staff, including through possible layoffs, by reassigning employees at its large customer sales unit who oversee relationships with major advertisers.”
They close by saying that, as of a year ago, Google had about 13,500 people devoted to the kind of sales work that can currently be performed by AI.
While many sources flat out say that Google “plans” to lay of 30,000 people, Ars Technica says that these “13,500 people aren’t necessarily all going to be affected, and those who are won’t necessarily be laid off—they could be reassigned to other areas in Google.”
This CNBC website, for instance, proclaims in their headline that Google is likely to lay off 30,000 employees, then in their text they tone it down by saying, “Google is considering a substantial workforce reduction, potentially affecting up to 30,000 employees.”
So which is it?
I understand that we live in the age of clicks, but of all sources, news sites should be responsible for accurate reporting.
Less than a year since ChatGPT hit the big leagues, it was already able to generate top-quality writing in seconds. 
On top of the challenges of SEO, attention spans have been shrinking for a while now.
The third trend that interests me is the fact that AI has become the new gatekeeper of all content.
There is another side effect of GenAI that many have yet to grasp: the ability to use GenAI to programmatically generate more GenAI.
Content marketing is at a crossroads. 
