It has happened to me several times to see advertisements for software to write posts in a semi-automatic way. Today I decided to test Article Forge; I report here my first impressions.
The home page is pretty clean, even if it doesn’t contain a lot of information. Unlike other similar sites there are no clear examples of the type of texts that can be written. I check the price immediately. There are two options, one with annual payment and one with monthly payment:
Doing the math, the lowest price is
13$*12=156$ a year, or $ 27 a month. However, there is the possibility to try Article Forge for free for 5 days. But you have to indicate a payment method.
I admit that this aspect bothered me. Payment is automatic when the trial days have expired. Before proceeding I wanted to read some other opinions, but unfortunately there is not much on the net. There are only two well-written reviews:
- Article Forge Review PROS & CONS (2022) Just Another Website Content Writer? di Kasa Review
- Article Forge Review di Terry L. Cooper
I decided to try Article Forge for 5 days.
I begin by studying the graphical interface. On the left are the sections available on the site:
- New Article
- My Articles
- Post Scheduler
- Usage Information
- API Information
- Help Centers
I’m interested in understanding how to create a new post, what options there are and, above all, the quality of the final result.
The steps to take are simple: I choose the title of the piece, the language, the number of words to write, then I leave it to the AI to write the piece.
By following the on-screen commands, I can choose between 7 languages:
As for the length of the piece, I can choose between:
- 50 words (very short)
- 250 words (short)
- 500 words (medium)
- 750 words (long)
- 1500 words (very long)
There is the possibility to automatically create sections in the piece (but this only applies to posts of at least 500 words). As for the title of the paragraphs, I can let the AI take care of it or write them myself.
Alternatively, it is also possible to indicate the keywords.
Finally, you can add videos, images, customize the links in the text, generate multiple texts on the same topic and even publish the result on wordpress.
How to use AI to write text
Well, I start playing with this software. And since Italian is available, I immediately try to have a piece written in Italian.
I choose Cose da fare a Milano (What To Do In Milan) as the title, set the length to 250 words and here is the result
The result is passable. The second paragraph is poorly written, but the whole is acceptable overall. I’ve seen human copywriters write much worse.
I try to do the same thing with English, with the title What To Do In Milan. Again I have to wait a minute or two
This time the result is better, both in terms of grammar and content quality.
Now I try other titles, always in English. First I enter The best rock songs as a title and set as an option the possibility of having an image and a video.
The result is a piece with two YouTube videos, with a list of songs. From a grammatical point of view it seems quite correct, but the content is not that great. It almost seems like a list of songs and phrases picked at random from other articles.
Then I try with The Cusbism, always with images, videos and paragraphs.
In this case, Article Forge offers a better piece. In the head he inserts an image while at the end he adds a video on the history of Cubism. The AI has decided to present some information on various Cubist artists, and the result is good.
Finally I want to try something absurd. In Italian I use Storia di un brigante e di un canarino (Story of a brigand and a canary) as input and I set the length to 250 words.
In English it becomes Story of a brigand and a canary. After a couple of minutes I get this warning:
The result is truly meaningless.
So, to conclude. The results of this test are mixed. I think Article Forge can be a useful tool for writing pieces on very common topics. But for somewhat strange issues, or for opinion articles, well, in my opinion it’s best to let it go.
Is it worth the price? I honestly don’t know what to answer. Personally, I wasn’t thrilled with the results. But it is also true that in general what I write is quite niche. It would be interesting to try to see in the field, with a longer test over time, what kinds of results can be obtained.