Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
And we want to know more and more: Is he a human or an artificial intelligence? Is that photo real? That song? This video? Substack publishing platform tries to add some clarity through Partnership with AI detector Pangram To help readers understand how AI is present in all the content they consume on the publishing platform.
The tool is available to all Substack subscribers and works on posts, notes, comments and responses published on or after Tuesday that are longer than 100 words. Substack said that content creators will also be able to scan their drafts with Pangram to see how spotty the AI is and can also submit a statement to explain their writing process.
Web and iOS users can try out Pangram’s features now, and Android users will have access to them later.
In a blog post Announcing the new feature, Chris Best, co-founder and CEO of Substack, said his platform aims to avoid “claudefishing,” i.e. when a reader consumes content without realizing it was generated by artificial intelligence.
“When readers question whether what they read is real, it undermines confidence in authorship and threatens the livelihoods of writers — including those who thoughtfully use AI tools to produce works they believe in,” Best wrote.
From CNET: AI Slop is destroying the Internet. These are the people fighting to save her
Pangram estimates approximately that 40% of texts on some social media platforms It is the gene of artificial intelligenceIt has an impressive two-thirds rating on the networking site LinkedIn. And AI growth agency Graphite recently published a white paper stating that AI is now writing Many articles on the Internet are like humans. A Gartner survey found that 68% of people They wonder if the content they see is realHalf of consumers prefer brands that do not use artificial intelligence in their marketing.
Tested it myself
I recently joined Substack, but wanted to check out the new Pangram AI detection tool. The tool works on posts, notes, comments, and replies, so I decided to use posts as a temporary testing lab. First, I logged into my Substack account and published a new post. I copied and pasted text from an article I had written – myself, without using AI – into a Substack post and published it.
After a few minutes, I clicked on the live post. In the three-dot menu at the top right of the page, I saw the “Scan for AI text” option. I clicked on it, and a small window popped up neatly declaring that my article was “entirely human-written.” A line appears below this indicating the percentage of text that uses AI, AI-assisted, or human text.
Next, I took the same post but added a long, completely AI-generated paragraph that I pasted from ChatGPT. When I ran the “Scan for AI Text,” the tool reported that 10% of the article was written by AI, which is an almost accurate estimate.
I then published a post that was entirely AI generated, and the scan correctly identified it as 100% AI generated.
In another test, I added a few AI-generated sentences to a post that was otherwise purely human content, and the scan flagged it as 100% human-generated. Close, but not quite right. So, the Pangram tool worked well in my initial experiment.
Although I was fairly satisfied with how well Substack’s AI detection system worked in my content, I was disappointed to learn that one of my favorite podcasters had gone full AI in his latest post. However, other writers I reviewed used entirely human-generated text.
What about false positives?
In his announcement, Best acknowledged that the AI detection tool was “not perfect,” but cited independent research that found that Pangram was 97.5% accuracy In full AI generated text detection. The University of Maryland found that Pangram was 99.3% accurate in identifying “humanized AI-generated text,” which basically means that a human has rewritten or reworded the AI content.
Substack authors who don’t use AI are understandably concerned about being incorrectly labeled as AI-powered authors. As a result, the risk of false positives is a major problem. Substack said creators can screen their content to see if it will be labeled as containing AI while it is still in draft form. If Pangram incorrectly defines the AI script, the creator can report the error to Substack and remove the faulty scan.
To promote greater transparency, publishers can now include the phrase “How do I do this” if they want readers to know how much, if any, AI they have used in their content. Best said, Pangram cannot detect whether AI has been used to gather information or as a source for written articles.
“We are not against people using AI to help them in their work, and we believe that people should have the freedom to choose the tools they use to express themselves,” Best said. “But people should know what they’re getting.”
See full bio
