AI is here for Serif fonts

With strong public opposition to the seemingly ubiquitous use of artificial intelligence, collective attempts to eliminate and reject signs of artificial intelligence use continue.

I am dismayed that one of the first victims is a dash — by the way, it is a great, very human symbol form! And there is also the “triple rule”, which is designed to appear rhythmic but often predictable, vulgar and old. And, of course, there's the awkward syntax structure of "not X, but Y."

At present, certain fonts and words (especially liners) appear to be defining (and abandoning) artificial intelligence, whether in actual software or in the design template for vibrating coding. Some people call it "tasteslop," which is the result of an effort to make the creative artificial intelligence design appear to be complex or outstanding.

Keya Vadgama, a writer, designer and font practitioner in the San Francisco Bay region, described this shift from a smoother and more visible computer-based font as a “lineline renewal”. Vadgama, in a recent newsletter published on her Substack, stated that it was intended to give the company more “personality and warmth”.

She wrote, "It's not hard to understand why AIS locals are particularly attracted to liner fonts: AIS is essentially cold and unconvinced." But the real humans used our product! We swear!”

"The line is based on calligraphy," Wadgama told The Connect magazine. “It means a very human, fluent type of font. "Vadgama notes Anthropic Yes. Claude. Default to use a liner. Other AI companies - Runway, Perplexity, Manus - have adopted similar fonts in their user experience and brand.

In an interview with Connect magazine, Jesse Dwyer, Chief Communications Officer, said, "Why don't we use human design? Perplexity is designed for humans."

According to Vadgama, the line is used not only for beauty but also to build trust between users and brands. Some font choices show confidence even at the psychological level of some pre-consciousness. Lineless fonts (Arials, Calibiris, Helveticas) are too clean and computerized. Good old Times New Roman, similar layout design, would feel a little more modest. Recently, Vadgama is working with a (lost) artificial intelligence start-up company, which favours liner text. "A large part of it," she said, "Yes, "How do we position ourselves to keep people from being afraid of us?"

Lines can help build that belief, or at least an illusion. Times New Roman himself was commissioned by the British Times in the 1930s. The font has an authoritative weight. Books and newspapers are printed in it. It was almost standardized in the decades before screen reading appeared. Perhaps the most famous is the Encyclopedia of the British — the authoritative framework of human knowledge, to say the least, before the advent of the World Wide Web — in the context of The Times.

“In the general public, the liner has an academic meaning,” said Ali S. Qadeer, Director of Plan Design, Ontario Academy of Arts and Design, Toronto. "Claude is funny. It mirrors the pages with a slightly brown background. It's kind of like reading print. Print is more deeply linked to trust.”

According to The New York Times, even the United States Department of State has re-used Times New Roman after Secretary of State Marco Luvio denounced Calibri as “irregular” and linked the Department of State to a number of broader DeI initiatives of the Biden era in unlined fonts.

Qadeer and Vadgama both see a line-line trend as a rebuttal of the perceived (in fact, literally) lack of soul of artificial intelligence and the widespread public scepticism about this technology. They're not the only ones. In addition to “low taste” statements, there are criticisms of the serialization of artificial intelligence aesthetics as “universal” and “very ugly”.

“If half of the people on the Internet begin to develop `unique and interesting' fonts without knowledge of the basics of design, then these fonts are linked to slop,” an X-user wrote in a post on artificially intelligent web design.

Humanization

A few decades back, computer-based designs were crude and ugly: slime green fonts fluorescent on heavy computer terminals. Think of those old IBM terminal fonts. When you think of them, your eyes get nervous. For Kadir, who taught the history of printing, this change was part of a deliberate rebranding. "I'm 100% convinced it's about softening people's emotions," he said. “This is a response to massive social criticism. The tedious appearance of technology that has dominated the past 20 years has become increasingly negative.”

Not everyone is so desperate. While the designer and founder, Zhang Yiyi, described the liner transition as “the curse” on X, he did not believe that it boded for anything particularly sneaky or dangerous. "Someone in these laboratories is trying to make these models design, he says. "It's very practical."

Chang compares the current aesthetic maturity of artificial intelligence to youth downloading and trying different fonts. (What age-specific computer natives have not downloaded Iron Maiden or South Park TTF font packages to do it?

He compared this emerging style to “quality mediocrity”, a concept proposed by the blogger Venkatesh Rao to describe a popular man-made luxury. ("High-quality mediocre," Rao wrote, "is the best bottle of wine in the olive park." Of course, artificially intelligent design choices are largely self-replicated, and as models are trained in the output of other models, the line will produce more liners.

Good Claude himself confirms that. In suggesting an explanation for its shift to liner fonts, chat robots cite issues such as trust, authority and “literacy”, while also explaining the “large borrowing” and network-based mentality. It also seems to recognize that the design choices of a smooth, complex and stylish nature are an attack: "Clean lines, fluid animations, sure layouts are communicating `the system knows what it is doing'. This aesthetic is contrary to an accurate psychological model of artificial intelligence.”

Wade Gamma agrees. “I do think that it is a little dishonest to try to use a line to show that `we are not one of those terrible artificial intelligence companies', but in fact, you are still,” she said. "You can use Comic Sans if you want. That doesn't prevent you from still being an artificial intelligence company."