HomeIndustriesMeet Joelle Pineau: She's shaping AI because the world grapples with its...

Meet Joelle Pineau: She's shaping AI because the world grapples with its potential

Joelle Pineau first encountered the concept of bias in artificial intelligence 1 / 4 century ago as a student on the University of Waterloo, Ontario. She had been recruited for a project on the university to assist train a voice recognition system for helicopter pilots.

To construct an unbiased system, they wanted female pilots to coach within the technology but couldn't find one available. So it was Pineau, an engineering graduate, sitting within the cockpit next to a pilot, whose voice was recorded to coach the system to the extent of stress a pilot might normally experience.

“I accepted the challenge,” she remembers. “It was a really rigorous scientific experience, this experience of mentioning biases in your data and finding an answer to them. Not that we had enough samples to make a really convincing argument, but we could not less than have a measure of bias and system performance to report on the gap.”

Pineau, now 49, speaks virtually from a room in Meta's New York offices called Reproducible Research. Today she is vice chairman of AI research at Meta, the technology company that owns Facebook, Instagram and WhatsApp.

Meta is a frontrunner in the event of AI. Its Fundamental AI Research (Fair) Laboratory conducts Meta's long-term research and goals to attain scientific breakthroughs in technology that can be implemented into its products across platforms.

Chief Executive Mark Zuckerberg has laid out his vision for constructing artificial general intelligence – whereby systems have a level of intelligence comparable to that of most humans. Meta says it wants to construct this on open source and make the technology easily accessible to developers and academics, moderately than hiding behind closed proprietary systems.

Llama, its series of enormous open-source language models, competes with closed models from competitors resembling OpenAI, Google and Anthropic. It has been utilized in Meta AI, an interface that enables users to talk in WhatsApp and Facebook, in addition to image generation.

Meta's advanced research is conducted through Fair, where Pineau leads around 1,000 people across 10 locations. “It’s an enormous responsibility,” she says. “Sometimes I see things in our research lab, after which it only takes a number of months for them to be within the hands of thousands and thousands, if not billions of individuals.”


Where did the trail come from? Budding musician and mathematician helping to shape among the world's strongest technologies?

Pineau's fascination with complex topics grew out of a love of mathematics “from a young age,” she says. “I remember being fascinated by basic mathematical concepts resembling power laws and the logarithmic base.” His keen interest in music – playing the piano and viola in orchestras – ultimately remained a hobby.

Pineau: As Vice President of AI Research at Meta, she leads the advanced research conducted within the Fundamental AI Research (Fair) lab © Kimberly M Wang

At the age of 19, Pineau first worked for Canada's Department of Natural Resources, constructing models of solar energy in fish farms. She later worked in a hearth research laboratory in her hometown of Ottawa. She has a doctorate in robotics from Carnegie Mellon University, where she built driving robots that might detect and avoid obstacles, in addition to assistants for nursing homes.

That's when she discovered AI and machine learning, developing algorithms and conversational chatbots – foundations for the advanced technology behind Meta AI and OpenAI's Microsoft-backed ChatGPT bot. “It was very primitive back then, within the late Nineteen Nineties, but it surely was fun.”

When he was teaching computer science at McGill University in Montreal, “there got here a time when it was pretty obvious that numerous the most important innovations in AI were going to occur in industry,” says Pineau. So in 2017, she joined Meta, then often known as Facebook, to open the Fair lab in Canada, alongside its existing labs in Paris and New York.

“I used to be interested in what you can do in research while you had investments on the industry level,” she says. “Meta was the one (company) committed to open science and open research. That’s why I didn’t hassle interviewing anywhere else.”

Pineau didn't got down to change into a frontrunner, she says, acknowledging that “it's just a little difficult to seek out individuals who have each the power to do research and the will to tackle more management responsibilities.” But she is now a part of a management team that consists more of ladies than men.

“I didn’t hire most of those people for (Fair), but I created opportunities and supported their roles,” she says. Then she adds: “As a female leader, additionally it is easier to draw other women. . . So that’s one other necessary a part of the equation as well.”

During her university profession, Pineau was one in every of 15 women in a cohort of 75 at Waterloo and one in every of six of 20 in robotics at Carnegie Mellon. Today, the number of ladies entering the technology sector stays low and, in 2019, also in skilled services corporations Accenture found that half left the corporate by age 35.

“Everything I say must be filtered through the filter of survivor bias. . . I’m still energetic on this field, probably because I’ve actually had a variety of quite positive experiences,” she says. Pineau points to the support she has received over time from a variety of generous colleagues and contacts. But she was also “very aware” of the imbalance between men and ladies on this area.

However, Pineau believes that science has made “remarkable progress” and that measures resembling allowing students to specialize or change their subject during their studies could help improve gender balance.

She remains to be affiliated with McGill University and lives in Montreal, where playing music, running and spending time with friends outside of the tech world gives me “a really different perspective” than her colleagues in California.

This distance and her long experience in AI work have given her beneficial perspectives because the world grapples with hype and concern in regards to the potential of this powerful technology.

“No one can really predict the long run, but there is unquestionably a moment (happening now),” she argues. “Since the launch of ChatGPT, for the primary time anyone, anywhere, can experience AI in a very different way. Before that. . . it was form of invisible.”

For Fair, there are still “numerous open research issues” and, more broadly at Meta, questions on the product and regulatory side.

“There is numerous scrutiny, but in addition numerous really good and necessary questions that we as a society need to choose,” Pineau says. “How will we position ourselves? There is not any right or incorrect answer. In what ways do we wish to work together to create an environment that promotes innovation, keeps people protected and respects the work of creators?

“We have to seek out the best ways to tie all of it together.”

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Must Read