Toronto startup Logojoy is giving potential customers a peek into the algorithm it uses to produce artificial intelligence (AI)-designed logos.
In a Jan. 24 blog post, Logojoy co-founder and CTO Rares Crisan explains that while logo design is subjective, the company’s platform studies how users interact with its design tools, and learns from each one.
“The goal of Logojoy’s programming is to understand what common features are in the logos our users love,” he writes.
With each new customer – according to its website, the company has served more than 1.4 million users and counting since being founded last July – it better recognizes which fonts and colours users prefer, which moods they align with, and the optimal sizing and spacing of visual elements.
“In other words, this data looks at how millions of users use an online logo maker,” Crisan writes.
Logojoy presented its information, naturally enough, in infographic form, which the company promises to update as its platform collects more data in the future.
“It may surprise people, as it did myself, to learn that there are a lot of mathematical patterns buried in these seemingly simple designs,” Crisan said in an emailed statement. “Machine learning sometimes appears like it needs to be this incredibly advanced decision maker that mirrors human thinking right away. In reality, it’s just about laying the foundation for gathering and inputting the right data.”