Skip to Content

Sony Research’s Parisian Play Station

Why would the multinational media conglomerate drop millions on emerging technologies that have no practical application? Because the future is closer than you think.
January 10, 2005

Last year, a video circulated around the Internet depicting a delighted German Shepard sniffing at a fine cut of beef on the floor while a Sony AIBO robotic dog watched curiously.

It was a cute scene, until the AIBO became a bit too interested in the steak. Without warning, the real dog viciously attacked the AIBO, while off-screen human witnesses shrieked in fear.

Who pitted the canine against the computer? Sony researchers themselves, in collaboration with Hungarian animal behavior experts. And that’s just one of the myriad experiments conducted by Sony Computer Science Laboratory (CSL) Paris, a commercial laboratory where research and development don’t always go hand in hand.

“It’s a scientific lab, but not all innovation is based on science and not all science leads to innovation,” says laboratory director Luc Steels.

Founded in 1996, the Paris operation is Sony’s only computer science laboratory outside of Tokyo. It’s a tight ship with six researchers tucked away in the city’s Latin Quarter. Housed in a classic Parisian building adjacent to the buzzing campuses of the University of Paris-Sorbonne and Ecole Normale Superieure, Sony CSL Paris feels more like an engineering graduate school than a key research arm of a multi-billion dollar consumer technology and media company.

And that’s no accident. Not only does the laboratory enlist Ph.D. students to assist with research, Sony CSL Paris relies heavily on the academic credo of publish or perish.

“The people who work here have academic reputations,” Steel says. “Our researchers present at workshops and publish in academic journals. They play that game and seek recognition and respect in that context.”

If that’s true, Steels is an all-star academic player turned coach. In the 1980s and 1990s, he ran the University of Brussels’ Artificial Intelligence Laboratory and conducted groundbreaking experiments on bottom-up AI. Specifically, Steels showed how robotic agents could evolve their own language by sharing with one another their observations of the world. The research spawned emotional speech synthesis technology that was eventually integrated into Sony’s new humanoid robot QRIO.

Considering Sony’s quest to put an entertainment robot in every home, it’s no surprise that they’re funding far-out experiments in robotics. And the media company would almost certainly be considered remiss if its research staff wasn’t designing novel search engines for digital music libraries, one of Sony CSL Paris’s projects getting lots of buzz back at headquarters.

However, the path from the laboratory’s computational neuroscience wing to Circuit City is a bit less direct.

“For a long time, I was puzzled why Sony would sink so much money into things that don’t have much direct application,” says Emory University professor Philippe Rochat, a well-respected developmental psychologist who spoke at a Sony Research-sponsored conference in Tuscany. “(It’s) probably for prestige and advertising purposes, but nevertheless, it’s unique and great that a big corporation will put some of their profit to identify people they consider the smartest in the field and let them come up with new ideas.”

Sony CSL was founded in 1988 by Toshi T. Doi, a computer scientist who was instrumental in the development of the compact disc player. Doi hired Keio University professor Mario Tokoro, now president of Sony CSL, to run the lab. Inspiration came from the Xerox Palo Alto Research Center, the quintessential corporate computer science laboratory known for its unbridled creativity.

Indeed, Doi’s friend and former PARC visionary Alan Kay recently spoke at Sony CSL Paris during their bi-annual open house in October. The event takes place in Tokyo on alternating years. That’s the day when “publish or perish” becomes “demo or die.”

“Thousands of people from all corners inside Sony come to watch,” Steels says. “But they come almost as scientific colleagues to figure out what you’ve really done and the potential of the work for themselves.”

The seed for the Paris lab was planted nearly a decade ago after Steels spent a few months as a visiting researcher at Sony in Tokyo. He was impressed with Tokoro and Doi’s vision and the company’s research competence as embodied by the robotic creature that would evolve into AIBO.

“The first prototype looked awful, but it was an impressive step toward bringing artificial intelligence into reality,” Steels says.

After the offer came to establish a Sony Laboratory in Europe, Steels quickly pegged Paris as the ideal location. First of all, he explains, there’s a high concentration of scientific activity within a bustling metropolis that attracts potential hires as well as a steady stream of visitors from academia. Secondly, he says, Paris has a cultural richness that’s a key ingredient in the laboratory’s multidisciplinary approach.

Researcher Atau Tanaka couldn’t agree more. He melds mobile technology with peer-to-peer networking to develop musical systems that transform listening into interactive social experiences. Tanaka’s work has garnered the attention of avant-garde artists and hardcore computer scientist alike. And if Sony someday turns Tanaka’s pet project into the next generation of Walkman, that’d be OK too.

“For me, the laboratory is a bridge between my scientific and artistic activities,” he says. “It provides the ultimate test to see if my personal visions can be validated in a truly scientific way.”

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

The problem with plug-in hybrids? Their drivers.

Plug-in hybrids are often sold as a transition to EVs, but new data from Europe shows we’re still underestimating the emissions they produce.

Google DeepMind’s new generative model makes Super Mario–like games from scratch

Genie learns how to control games by watching hours and hours of video. It could help train next-gen robots too.

How scientists traced a mysterious covid case back to six toilets

When wastewater surveillance turns into a hunt for a single infected individual, the ethics get tricky.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.