[ad_1]
On the surface, Chat GPT looks like a great solution for people to be able to quickly generate and publish content for their blog that will rank highly on Google. There are countless YouTube videos explaining how to do this.
AI content is bad for SEO and ChatGPT should not be used to generate articles for blogs. AI content lacks original thought and will not highlighting a writer’s experience. The content generated will fail Google’s E.A.T. standards and is unlikely to rank well in Google Search Results for SEO.
However, there are many use cases where AI can improve a writer’s productivity.
Specialized AI copywriting tools continue to improve over time and are capable of completing baseline research that writers would have had to spend hours completing prior to even beginning to write an article.
Let’s dive in!
One of the primary designs of Chat GPT is for content creation. It has the ability to ingest massive amounts of information, re-sort, organize, and summarize it for humans to understand. It’s contextually aware of the questions being asked of it allowing people to ask follow-up questions.
Prompt Chat GPT to create a 800-word blog post about Canadian Bacon and it will blast out an article in under a minute. Chat GPT-4, the latest version of Chat GPT can do this with ease.
Other AI content generation tools have popped up recently that allow you to create photorealistic or stylized images such as midjourney.com or OpenAI DallE-2.
These tools are incredibly easy to use, hyper-efficient, and can be massive time savers. So, why are we hesitant to jump on board with Chat GPT-generated articles for SEO?
By nature of how Chat GPT works, it assimilates existing information. Mostly information prior to September 2021, but there are alternatives that are trained on current events such as Google Bard and Bing Chat.
Because Chat GPT is assimilating existing information, and often slightly outdated information by nature it is almost guaranteed to not be original content.
If we examine the criteria that Google lays out for content that will rank well, we find the following excerpt from Google’s own website:
You can read the full article here:
Many SEO experts summarize Google’s content criteria with the acronym E.A.T. which stands for, Expertise Authoritativeness and Trustworthiness.
Let’s break down each component of these critical SEO components.
- Expertise — Demonstration of in-depth knowledge on the specific subject matter. This includes providing comprehensive and insightful information for the reader.
- Authoritativeness — Recognition as being an expert in the related field. Established credibility among the online community with backlinks being one of the primary indicators of how valuable the content is.
- Trustworthiness — Delivering accurate, honest information and clearly disclosing any potential conflicts of interest.
By applying Google’s criteria to content generated by ChatGPT it is unlikely that it will demonstrate expertise in a field and at worst, ChatGPT hallucinations could provide readers with inaccurate information when an authoritative author would not.
In regard to investing, Warren Buffet said, “Price is what you pay. Value is what you get.” We believe that the same is true and will be even more true in the future as AI content generation becomes more pervasive.
According to websitesetup.org, there are over 1.7 billion websites with 6 billion indexed webpages in the world today.
It’s likely that this number will continue to grow and the rate at which it goes will speed up. As it becomes easier to generate more content it lowers the barrier of entry to maintaining and generating content for a website.
To stand out in an ever-growing crowd, content creators will have to create material that provides value to the readers. Generating value is incredibly difficult and is often a function of time, effort, and talent. Shortcutting the process and not putting in the work, or offloading the work to an AI will likely result in mediocre results.
Yes and No. There are new Chat GPT-powered tools that allow people to quickly and easily generate content. Need more words? Click the generate words button! It’s just that easy! Unfortunately, there’s no indication that having more words provides more value.
Google has indicated that article word count is not specifically a ranking factor. Though, many SEO experts have pointed out that longer articles have more keywords and give Google’s indexing algorithms a greater chance of determining the quality of content and how to rank it.
One of the leaders in AI-powered content creation is Jasper.AI which is specifically designed to help writers create better content that will rank on Google. It’s a paid service with mixed reviews.
To learn more about the product and see it in action, here is an example of Jasper generating technical content for a blog post.
We believe that there is value in tools like this not necessarily to generate content for you, but to act as an informational tool to guide which keywords and topics will be key signals for Google SEO.
At this time, Google does not ban or discourage AI-generated content. It is unclear if they will do so or not in the future. As Google works hard at improving Bard, their own AI Chatbot rival to Chat GPT it would seem odd that they would discourage people from using it.
It’s important to note that Google also does not have to specifically ban AI-generated content for it to not rank well in their search results. If the content provides little additional value, is poorly written, and does not contain original thoughts Google can rank your site poorly without using AI as a specific ranking factor.
We have not heard of any specific confirmation that Google is searching for and detecting AI-generated content in their search results. However, per Statista, Google had over 190,000 global employees at the end of 2022.
Knowing this and that AI is one of Google’s primary focal points, it would be shocking if they did not have a group or multiple groups of people creating a solution to detect AI content.
We tested the AI Content Detector available through writer.com with a paragraph of text created by Chat GPT-4. The detector failed, stating that it was 97% human-generated content. It’s not surprising that it failed. AI is designed to create text like a human would, and it will continue to improve at it.
When we tested writer.com on a block of text regarding the same topic that was generated by Chat GPT 3.5, it correctly produced the result of 34% Human-Generated Content.
Another topic that Google, Microsoft, and OpenAI are likely trying to figure out is how to feed fresh content into their AI platforms without creating a feedback loop.
We could see a couple of different ways that this could go. Either the tech companies create technology that can detect AI-generated content. Some of these solutions already exist, but over time will become an arms war of AI generation technology that doesn’t want to be detected and the AI-generated content detectors. It will likely become its own industry.
The second option would be to tag content that has been AI-generated. This could be similar to the way that Nofollow links currently work today. Links are the currency of the internet, and by tagging them nofollow they are excluded from being taken into consideration for page ranking by Google.
It’s possible that content could be tagged as AI-generated, or placed with a NoTraining tag so that the content is excluded from large language model training. It would be an easy solution and give content creators more control over what happens to their data once it is published to the World Wide Web.
Because Chat GPT and its alternatives are so new, the web search industry has not had time to adjust and determine what that best approach will be.
In 2017, Facebook experimented with and shut down an AI system that created its own language that humans couldn’t understand. We don’t bring this up to scare anybody, but it does illustrate that researchers may not even know what will happen if AI is trained on information that is generated by other forms of AI.
Not to be overly gloomy on the prospects of using AI to improve your SEO rankings, there are some very useful and creative ways that AI can be leveraged to help you create better content. Here are a few to consider:
Content for highly competitive keywords will need to exceed the value of the content that already exists for those keywords. AI tools can summarize the competition and identify trends in keywords that Google will expect to see. This information can be used as a starting to point to create baseline content that you can add original thoughts to and build upon to add more value.
Chat GPT is a very useful tool for organizing thoughts. You could either take a bunch of random ideas related to a blog article that you are thinking of creating, or you could go through your own content research process, copy and paste it into Chat GPT and let it create a content outline for you.
There are times when finding the right words just doesn’t happen right away. Or maybe you write a paragraph, re-read it and realize that it sounds terrible. Try copying and pasting it into Chat GPT to suggest ways to reword it. Include additional information in the prompt to ask it to make it more concise, longer, more professional, or re-write it at an 8th-grade reading level.
If you think of Chat GPT similarly to being your intern, personal assistant or employee there are many times when we’ve asked employees to write a draft of a document knowing that we were going to put it into our own words. Chat GPT generated content could be used as a starting point for you to build off of or for you to later re-write.
Microsoft is positioning Microsoft Copilot 365 as a way to generate draft documents directly in Microsoft Word, powered by Chat GPT-4. Google is working on a similar solution for their Google Workplace suite of business applications and will probably become the new normal for writing.
Trust but verify is our motto when using Chat GPT or more up-to-date options like Google Bard. Since the release of Chat GPT-4, the results that the chatbot provides seem to be more accurate more often. However, that does not mean that we don’t attempt to validate the advice it is giving through other sources.
Chat GPT and Bard are very good at providing background information where it is less critical that the information is 100% accurate. Being able to explore sub-topics by asking questions or for clarification are fantastic use cases. We have even used it to copy and paste pages of technical documents into it for Chat GPT to summarize and explain in other terms.
Chat GPT is already changing how people work and live. The internet will never be the same after its release. The new technology makes it easier and less time-consuming than ever before to generate new content for blogs and websites.
However, the content being generated is based on the average of information that is already readily available. This makes it unlikely to rank well on Google for the purposes of SEO. Original content that provides readers with the most value will continue to make it to the top of the page rankings and high-value content will still require a high level of effort to create.
That being said, there are many ways that Chat GPT can help content creators streamline their work. From aiding in the research process to helping people get past writer’s block, there are many benefits and exciting use cases.
[ad_2]
Source link