TikTok users love to watch this guy’s drawing robots screw up

Few TikTok users will have escaped the idiosyncratic videos of @robotsdraw on their For You feeds in recent weeks.

The account, in which a robotic arm draws with amazing accuracy on a piece of paper before invariably screwing it all up, has become enormously popular, racking up more than 15 million likes and a quarter million followers on the app, the majority of which have come in the last two months.

This success comes as a surprise to the human being behind the account, San Francisco Bay area entrepreneur and angel investor Joshua Schachter. After all, it began simply as a way for him to show off his favorite hobby: tinkering with pen plotter robots.

If Schachter’s name sounds familiar, you’re likely showing your age: The 48-year-old is the founder of early web social-bookmarking service de.licio.us, which allowed users to save their favorite websites and share them with friends. Since selling de.licio.us to Yahoo! in 2005, Schachter has spent his time funding startups, in addition to acting as a consultant for HBO series Silicon Valley and an advisor to Walmart Labs.

Schachter’s interest in pen plotters grew out of an obsession with discarded tech. “I used to go on eBay and buy interesting failed gadgets of old that I could get for less than $20,” he says. “I’ve always been interested in evolutionary pathways that didn’t work out.”

A particular interest of Schacter’s was computer-controlled machining. “Seven or eight years ago, I built a whiteboard that could draw on itself,” he says. He later constructed a machine that deposited ink on paper. “I got interested in the algorithms behind this,” he says. “I learned how to do procedurally generated art — where the computer makes various decisions using all sorts of techniques.”

He moved on to collecting vintage pen plotter machines — essentially robotic arms that draw in response to commands sent to them by a computer. In September 2020, he posted his first video on TikTok, which captured a 1989 Hewlett-Packard DraftPro DXL producing a beautiful range of squiggly lines, then followed it with several dozen more using the same machine.

At first, the videos were snapshots of the machine’s broader movements — small snippets of the pen drawing one of hundreds of similar shapes, accompanied the hashtag #oddlysatisfying. But then Schachter began to pivot.

It started with failed attempts to try and get a paint extruder to work. Then, in late April, Schachter posted a video of that machine creating a square of dots in a seemingly random way, adding huge inefficiencies to what could be a simple process. The top comment on the video is “The most unsatisfying satisfying thing I’ve ever seen.”

“People really emotionally connected with it,” says Schachter. “They’re like: ‘He’s trying’” — “he” being the robot — “and stuff like that. There was an interesting connection.”

Finally, Schachter hit upon the premise that would take his account to the next level: robots failing at stuff. “I realized that people will sympathize,” he says. “The motions of a thing are relatable.”

A video posted May 5 of a robot drawing 44 perfect small circles in purple ink before fluffing the 45th and wildly scratching out the whole picture got 3.8 million views. @robotsdraw practically doubled its followers overnight, from 5,900 to 10,000. Comments came in, suggesting he introduce other mild errors, like misplacing a single circle — which Schachter was happy to comply with.

The reason @robotsdraw is popular is simple, says Jonathan Aitken, academic in robotics at the University of Sheffield in the U.K. “As humans we think that robots are infallible,” he says. “When we come to look at robots, any error is a deviation from our expectation and ultimately grabs our attention.”

The videos, with the rhythmic rat-a-tat of a robot dotting paper or a dry-erase marker squeaking across a whiteboard, also tap into the popularity of ASMR content online. “The shapes and sounds, and the scratching of the pen trigger different memories,” Schacter says, “which then connect to other things. I try to take advantage of it.”

In May, Schachter’s video of a plotter drawing a perfect circle and lines crossing through its middle, only to mess up on the final one, also garnered millions of views. “Now I have two narrative arcs I can play on,” Schachter recalls thinking: “A pattern damaged, and an emotion expressed. I started chipping away at both of them.”

Emotional connection

Schachter has kept the schtick going in the weeks since, spending around an hour a day conceptualizing and shooting videos using his plotters. The creator tries to read every comment posted on the videos — a challenge as the account has become more popular.

He’s also developed a kind of manifesto for @robotsdraw: “Intrinsic to the plotter is the position, speed, pen color, and sounds. Extrinsic is the pens themselves failing, ink failing, paper failing. Emotive: the ideas are angry, love, pausing, happy, hesitant, frustrated, different kinds of expectations and failures.”

It’s not coincidental that those emotions are many of the same ones that TikTok advises its top creators to target to make successful videos.

“I’m fascinated by TikTok,” says Schachter. “I’ve actually sent in my resume twice, but got the standard rejection.” He sees echoes of the social web he helped pioneer in TikTok, but improved by the switch to shortform video. “The narrative arcs, the standard formats are just now evolving, which means there’s a sudden turnover and democratization of who’s good at this.”

Schachter’s own success in finding out what’s good on TikTok has helped him grow the audience for @robotsdraw, 60 percent of whom are male and 55 percent of whom are based in the U.S. — and almost all of whom seem hooked on what the machine will get wrong next.

“They’re not friends with the robot,” Schachter says of his viewers. “They often are annoyed by it.”