D-Wave Systems Inc. co-founder and CTO Geordie Rose will be presenting a new AI venture called Kindred Systems, along with Kindred founder and CTO Suzanne Gildert, at the Creative Destruction Lab’s Machine Learning and the Market for Intelligence conference at the Rotman School of Business in Toronto on October 27.
D-Wave is a Burnaby, B.C.-based quantum computing company, which has made a computer that employs a quantum annealing technique, producing quantum-like effects.
Kindred Systems filed a patent in August relating to the creation of a robotic exoskeleton attached to an array of sensors, with the intention of teaching robots working remotely to mimic the exoskeleton wearer’s actions until the robots learn to reproduce the action.
Rose and Gildert will be elaborating on their plans on a panel called “What is Intelligence?” along with biologist Frans de Waal, author of “Are We Smart Enough to Know How Smart Animals Are?”
The patent filed by Kindred Systems is called “Facilitating device control” and spells out a method of directing a “device” to “take at least one autonomous action” after receiving instructions from sensor information which causes it to simulate an operator-triggered action.
A little bafflingly, the patent also mentions monkeys, which may explain the presence of animal expert Dr. de Waal on Rose and Gildert’s panel.
“In various embodiments, an operator may include a non-human animal such as a monkey, and the operator interface may be generally similar to the operator interface shown in FIG. 16 but re-sized to account for the differences between a human operator and a monkey operator,” says the patent.
Kindred’s bare-bones website states that “No monkeys were harmed in the training of our AI.”
The result looks like an attempt to implement robots using artificial intelligence to do repetitive labour in a workplace environment, whether warehouse or domestic.
The Kindred robot is presented as a 1.2-meter-tall humanoid, potentially covered in synthetic skin, rolling around on wheeled treads, having at least two arms with hands or grippers, and at least one camera on its head streaming high-definition video to its operator, with other sensors detecting a range of environmental information, including infrared and ultraviolet visuals, a radiation detector, GPS, and strain or touch sensors.
In an August blog post on Medium, Silicon Valley venture capital firm Data Collective described Kindred Systems’ exoskeleton as “AI-driven robotics so that one human worker can do the work of four.”
According to Data Collective, Kindred has raised over US$10 million in funding, along with Eleven Two Capital, also of Silicon Valley.
In a September article published by the World Economic Forum, Harvard University Economics professor Martin Feldstein predicts that introducing robots into the workforce will benefit American workers, who put in 30% more hours than German workers.
“Rapid technical change is not something new,” writes Feldstein. “We have experienced technological change that substitutes machines and computers for individual workers for many years. And yet, despite the ups and downs of the business cycle, the U.S. economy continues to return to full employment.”
This week, D-Wave “quietly” raised US$21 million, led by Fidelity Investments and Montreal public-sector pension investment manager PSP Investments.
Last week, at D-Wave’s inaugural users group conference in Santa Fe, New Mexico, the company announced its next-generation system, using a 2000-qubit processor, effectively doubling the number of qubits capacity of the previous generation D-Wave 2X system, which the company sold to Google and NASA in 2013 for $15 million.
While highly secretive, Kindred apparently has about 25 employees in Vancouver, including several who previously worked for D-Wave.
Last November, Gilderts was spotted at a ladies-only “Battle of the Bots” event during which she expressed a belief that robots are evolving Artificial Intelligence, and revealed that she had been working on a Machine Learning system for robots to learn behaviours through data mining and to recognize patterns for decision-making.
More info should be forthcoming in Toronto later this month.