Higher AI models should be able to generate better, relevant and more contextual content
under review
Hans David
This is the first sentence produced by the Sonnet 3.5 model for a post titled "Prepping for 2025":
Are you ready for the unpredictable challenges ahead? Prepping for 2025 is more important than ever. The Texas electric grid failures remind us how fast our comforts can disappear.
The AI model instantly jumps to provide some random facts and make the content looks weird and awkward.
Isn't the advanced models should be able to humanize their content so that we don't have to work twice as hard?
And here is another content generated by the advanced model:
"Climate change is making natural disasters worse.
Economic troubles are also on the horizon, making it crucial to be financially ready. Political events could lead to civil unrest, adding to the uncertainty. To stay safe, you need to learn many skills and build strong community ties.
By 2025, innovation will be key for businesses and individuals. Companies will have to match their plans with their goals and find funding for new projects. For preppers, this means keeping up with new tech and updating your plans.
Cybersecurity threats are growing, targeting businesses of all sizes. Use strong passwords, multi-factor authentication, and keep your software current. Regular security checks and teaching employees are essential for a strong defense against online threats."
Look how abysmal the quality is. It just brings up random sentences that do not flow with one another.
You promise a humanized blog content that is of high quality. I don't mind doing some editing but everytime I generate a content, I have to do super duper heavy editing by adding relevant facts and improve the quality of article flow. If you ask people to pay for your software, you should be able to deliver what you promise. So far, what your software produces have been mostly abysmal.
This has been an ongoing issues. Random facts should not be popping out of nowhere without a bridge sentences or paragraphs. Fix this please.
SEOWriting.AI
under review
SEOWriting.AI
Did you have the "Connect to Web" function enabled or not?
Hans David
SEOWriting.AI yes and the problem arises with this is that the software does not understand how to curate their findings in a contextual manner to provide a logical article flow. It just picks up one sentence from a link and then combine it with a random sentence from the next one and so on. What happens is an abysmal article that is filled with keywords but is not friendly to the human readers to read. You software should not just be able to fetch random sentences from links but also contextualized them so that it can produce at least a mediocre level article that we can edit a little. With the way it works right now, users have to heavily edit almost everything. I have an article that I have to re-write almost entirely because your software fails to do its job properly.
SEOWriting.AI
Hans David “not just be able to fetch random sentences from links” we analyze the entire articles from the links. Ok, thanks, we'll review the process and make improvements.
Hans David
SEOWriting.AI It does "analyze" but not in a proper manner. Back before there was AI, bloggers look for relevant information all over the web, curate them and then craft a content using the information and contextualized them to the article.
What your software does by "analyzing" is just simply visiting the links, look for information based on the keywords and then just rewrite an random sentence from each link. It does not curate and contextualize them all together to become readable for the human readers.
That is how I see it after analyzing the generated content. I activated the relevant link settings and once I see the produced content and visited the links found by your software, I see the pattern.
You must train your AI software to think and act like a human blogger would when he gathers information from the web.
SEOWriting.AI
Hans David We don't rewrite random sentences from each link. Instead, our software analyzes all the content and contextualizes it.
We will continue improving our software to think in a more nuanced and human-like way, ensuring the content is readable and meaningful for human readers. Thank you.
Hans David
SEOWriting.AI this is another example on how the AI clearly just taking random sentences from links it finds without sufficiently contextualize it or making it make sense for the human reader:
"While self-preservation is key, it must be balanced with community support. History shows that collaboration leads to better outcomes in emergencies. For example, the Swedish Civil Contingencies Agency's 2018 efforts demonstrated the power of collective responsibility. Such initiatives show that a community-focused approach strengthens disaster readiness, creating a resilient network.
Psychologically, prioritizing personal safety can lead to isolation, causing some to become lone wolves. On the other hand, community engagement boosts morale and reduces anxiety in challenges. As civil defense education grows in Estonia and the U.S., the focus on group preparedness builds a culture of resilience. It encourages individuals to seek partnerships for mutual support."
Why is Swedish Civil Contingiency Agency is in there? And then in the second paragraph for some reasons it talks about civil education in Estonia and US (two countries that are thousands of miles apart for crying out loud).
Read the two paragraphs and their relation and flow with each other just does not make any sense. Yeah they are keyword-rich for SEO bots but for human? I would be confused with what I am reading there.
SEOWriting.AI
Thank you for the examples. We will definitely improve on that.