‘AI Content Creation Tactics – Stop creating poor content for your website!
Google is not going to rank it well, even if it ever indexes it.
The world is full of people who are trying to to get around Google rules, guidelines and algorithms. In 2023 and 2024 we saw a plethora of so-called ‘SEO experts’ recommending the use of ChatGPT from OpenAI or Gemini from Google, Claude from Anthropic and other AI or machine learning tools to create content for websites. Jasper, Wordsmith, Articoolo, Word AI or Quillbot to name just a few of the writing platforms. Just give them some key words or phrases, add any prompts, then let them spit out hundreds or thousands of words or code – simple. Yes simple, but a minefield for you if you do not edit it sufficiently and maybe check on the source. Whilst the output from a prompt should be ‘original’ content, it is not always factually correct. And you could get answers that are simply wrong, ‘hallucinations’ is the technichal term for such false information.
Back to the Future Content Tactics
Years ago before these bots or tools, these sorts of people would scrape content from other people’s websites, posts or articles to create me too content for their own client websites. Nowadays they are giving these tools to non native speakers and saying use ChatGPT, Gemini or Jarvis to create content. The result is streams of poor quality, uninteresting content. It could make the dumb dumber, so be careful who you let use it to create content on behalf of your firm. Your reputation is critical to your success. And critical for ranking in Google results.
This sort of content was of little value way back in 2010 and is of even less interest in 2024, because Google knows how to handle all of this junk much better now. As the output of content from AI gets better, Google will still look to deliver the best answer to any given question or search query. That has not changed. Adding lots of poor quality content to your existing website will only do damage and your rankings will tank. Google is not simply looking for content that is focused on technical optimisation for a given keyphrase. It is looking to find the best answer based on ‘intent’ of the person doing the search query. Is the intent to get information, navigate to information or do a comparison or make a purchase, as in transactional. Which part of the funnel or buying process are they at now? You may want to create content at each point of the funnel to keep your readers engaged, through to a transaction. Our mantra is ‘If it is’nt worth writing yourself then it is’nt worth posting’.
AI – Good or Bad?
There is nothing wrong with AI as such. This is effectively a ‘tool’ that will allow us ideate or create content and do things faster. Think of the AI tools as being an Assistant for you and your team. You do need to get to understand how to use AI in the same way as you did learn how to use a browser to access information on the internet through Netscape Navigator or Internet Explorer or Google. Your competitors are probably doing this already, so do not get left behind. It is a bit like the industrial revolution and it is that much of an innovation.
So that is all good. However, many content creators would say that AI is bad because it is doing their job and perhaps doing it badly. Other people worry that AI or AGI is going to replace humans in the long run. Stephen Hawking warned about this in 2014 saying it ‘spelled the end of the human race’. Sam Altman, one of two co-founders at OpenAi who created ChatGPT was relieved of his job in November 2023, after they had become the most well known name in this space. And his co-founder Greg Brockman resigned when he heard that news. When Altman returned, Brockman did too, but then he took extended ‘leave’ from the company, until late 2024. Neither agreed with the direction of the board of directors, as the board said they wanted more controls put in place. They stated that they were concerned about how AI was being used by OpenAI. They then appointed an interim CEO Mira Murati. She left in September 2024 with many other key people, to create their own AI firms. So it is clear that there will be wholesale abuse of AI tools in 2024, 2025 and beyond. But who will be the ‘bad actors’? Murati says that AGI artificial general intelligence will be achieved before 2030. If you think that the advent of the Internet was a revolution, AGI would eclipse it by a long way.
Update – Altman and Brock were reinstated and OpenAI is back in business, after a tumultous few days. Now Sam is looking for trillions of dollars to push forward at OpenAI in the race to AGI Artificial General Intelligence. And the board, they got replaced or quit, citing a lack of oversight and security at the firm. Both Microsoft and Apple subsequently invested further in OpenAI. Yet both declined the offer of a seat on the oversight board. September 2024 saw a mass exodus of key people as Altman took more control over everything. And as of October 2024 Microsoft who owns 49% of the ‘for profit’ enterprise has stopped any further major investment in Open AI. At the same time they have acquhired Mustafa Suleyman and most of his Inflection team to work on AI at Microsoft. The plot thickens, as they seem to be competitors now.
Note: If you let someone create content on your behalf you are still responsible for what gets posted on your behalf. That includes website content, video and social media posts. Ignorance is no excuse in the eyes of the law.
Plan Your Content Creation
If you want to write content for your website, it needs to be something new or fresh or different and ideally of interest to anybody who finds it. If you simply write about something that has already been written about a million times or even rehash something that has already been written somewhere else, Google will not feel your content deserves to rank. The Helpful Content update from Google basically states the Google will rank content based on how helpful it is, rather than just simply being relevant to the search query. Warning: It is a long read.
Content Indexing V Content Ranking
Currently Google is crawling then discovering pages and not necessarily indexing these poor or repetitive posts. Or at the very least there is a much longer delay to indexing than there used to be. A big part of this is Google letting you know that your new content is of no great interest, so your posts are not ranking well. What Google would like to see is that any content that you write gets shared and ideally get good authority backlinks as well. This shows it is appreciated and has value.
Content marketing in 2025
This is going to require that you only create good content and ideally create great content. Rather than create ten blog posts or articles about a topic, why not invest the time to create one high quality article. Then go about getting people to share and link to and mention your article in other websites. Google would really like to see you get backlinks from authority websites in your own space. Buying links is not a good idea. Yes it works for now, but less and less so going forward. Unless a link is from some high authority website and one that is related to what you do.
Content Creation in 2025
If I want to create an article on SEO as in search engine optimisation techniques then I would ideally want to be getting backlinks from other websites that are recognised to be expert or authority websites in the whole area of search engine optimisation. If I simply write ‘a n other’ article on SEO, Google will have no great interest. Therefore, even when it does eventually get indexed, it will not rank very well. Check out how Google defines Helpful Content here. The Spam Core update of October 2023 saw Google reindex many pieces of content to ensure that they are delivering the best answer to any given question or query. It caused mayhem for so many websites and their rankings in search. And rightly so.
What is SEO? – start here if you want more information on this huge topic. Now this is a big article and we did not use AI to write it – for better or worse, you decide.
Glossary of AI Terms
Some of the terms you hear when it comes to AI – here is a simple rough one liner about some of them.
What is AI? – At a basic level, computers can be fed with data in dataset or language models or images, so that when you ask a question or enter a prompt, it will deliver an answer based on its large language model. It produces answers based on its knowledge base. Currently, the output from Artificial Intelligence differs from Human Intelligence, as you can take information and add or improvise based on your experience. So our answer to any given question should be less robotic, more creative, intuitive, and hopefully more useful than any AI answer. However in our case we too are limited by our knowledge and experience. So a complex question could be better answered by an AI that was fed a dataset that we do not have at our disposal. So AI is the science of making computers smarter, faster, more efficient or productive. IBM describes AI as ‘a technology that enables computers and machines to simulate human learning, comprehension, problem solving, decision making, creativity and autonomy’.
What is a GPT?
Chat GPT is a General Purpose Technology and also GPT is used to describe generative pre-trained transformer models that are neural network-based language prediction types that are built on the ‘transformer’ architecture, a machine learning framework than can encode and decode. They can take any NLP natural language queries, or prompts, and then they predict the best possible response based on their understanding of language. Many are multi modal, so we can transform the written word to audio or video. Most are not exact and will need to be refined and fact checked.
What is Gemini previously Bard? And how does it work? Gemini is a Google tool a multimodal large language AI model that you can use to explore creative ideas and explain things simply. It’s a Google AI experiment that can generate text, translate languages, write different kinds of creative content and more. Gemini 1.5 now underpins NotebookLM.
The simple answer
When you enter a prompt into Gemini, it replies with a response using the information it already knows or fetches from other sources, like other Google services.
The technical answer
Gemini is a large language model using LaMDA Language Model for Dialogue Applications.
The language model learns by “reading” trillions of words that help it pick up on patterns that make up language, which allows it to pick up on and reply to your questions in common language patterns
It’s always learning, which means it also learns from your prompts, responses and feedback.
It’s an experiment, and it will make mistakes.
What is Googles Gemini? – Gemini is Google’s most capable and general AI model to date, announced on December 7, 2023. It is a powerful multimodal model that can understand and generate text, translate languages, write different kinds of creative content, and answer your questions in an informative way. Gemini is built on Pathways, Google’s new AI infrastructure, which allows it to process information more efficiently and scale to larger models. See more at https://deepmind.google/technologies/gemini/#introduction
What is a Large Language Model? – This a dataset or corpus of information, a body of knowledge as such. When fed in to a tool like ChatGPT it can then create an answer to any given question or prompt. Any answer or outout is limited by the extent of its data or knowledge from the LLM it was fed initially. These LLM used to cost a fortune, but nowadays they are getting incredibly inexpensive to create. This is why we are seeing AI tools becomiong available to the general public for small money. Or for free, in which case you are being given the tool, so that you enter the data, through questions and prompts and refining the answers you get. You are essentially working for the platform.
What is Narrow AI? – Think ChatBot or Siri or Netflix or even Google or any ecommerce platform that recommend things that it thinks you will like based on your browsing history. It usually has one single function to deliver a suggestion based on its training. This narrow ai has been around for many years. And it is cheap to create. Nowadays you can use AI tools to create a narrow AI, such as your own ChatBot.
What is the T or Transformer in ChatGPT? – The transformer was added in 2017 and it understands context. It can take your prompt or question and deliver an answer. Before the advent of the transformer AI was more limited. The transformers today can vectorise text, images or sounds so that it can deliver answers in text, sound or image and video such as Dall-E. The transformer encodes everything in to numbers essentially, so that it can then relate anything and generate output based on its understanding of context, through Neural Networks.
What is Machine Learning? – It is just that, a machine that is learning from everything we put in to it. Every time you ask a question it is learning. It further learns when you ask another question based on the answer it gave to you. It keeps learning as you keep asking until you get an answer that you deem to be correct. Once it learns from the first encounter with you, it can deliver the best answer quicker. And if people ask more questions based on that answer it learns even more.
What is Deep Learning? – This is a more evolved or evolving variation on machine learning as it uses more layers to determine an answer to a given question.
What is a Neural Network? – This is where a comouter is trained to think like a human brain. It breaks down layers of processes through neurons or nodes to determine an answer to any given question.
What is AGI? Artificial General Intelligence is the concept that AI becomes better and faster than human intelligence. Could it happen, should it happen, has it happened, this is all conjecture and above my pay grade. Certainly it is way off topic for a simple post about using basic AI for content creation in websites. If you want to dig deeper, do a search for Mustafa Suleyman at Microsoft AI and previously founder at DeepMind and at Inflection. Maybe watch his interview with Steven Bartlett HERE to get a feel for how he sees things going.
For most people this is enough to give you an idea of how AI can generate content, in text, audio, video or photographic mediums. The Generative AI can generate original text or image based on your prompt. It is a transformer to that output should be ‘original’, in the sense that it is unique, but it pulls from data that it was fed, hence it is generative.
Hungry for more information about AI in general?
Try listening to Paul Roetzer and Mike Kaput at The AI Marketing Institute podcast. This is fantastic weekly podcast. And these guys also offer training at any level for the C Suite. Or read their book Marketing Artificial Intelligence: AI, Marketing, and the Future of Business.
Image: Sequoia Sonya Huang Generative AI Market Map shows the fast changing Generative AI landscape in late 2022 with many new players coming in to the market. Nowadays we can generate written, photographic and videographic content plus AI voice or audio content. You can now have a ‘voice’ read your book rather than paying for a human to do a voice over reading. However, would you want to listen to this? Engaging, entertaining content will still win the eyes and ears of your website visitors.
