2023 is a year that has been marked by technological leaps. It is not surprising to see the confluence of artificial intelligence (AI) and robotics which promises to reshape industries and redefine labor.
Today there are about 3,000,000 industrial robots around the globe, with roughly 400,000 new robots entering the market every year. And the global market is expected to reach almost $43.8 billion in revenue in 2023 with a 10% CAGAR growth through to 2029 in North America alone.
I had the privilege of speaking with Peter Chen, the Co-founder and CEO of Covariant to dive into the implications of AI-infused robotics in manufacturing, the critical challenges in labor shortages and supply chain disruptions and the transformative journey that Chen and Covariant are pioneering.
Chen’s Roots in OpenAI
Peter Chen’s journey in the AI landscape finds its roots at OpenAI, today’s darling of cutting-edge AI research. Chen and fellow Covariant Co-founder Pieter Abbeel and Rocky Duan were all with OpenAI at its early stages when the team comprised of 10 members. Their tenure extended until the organization had grown to approximately 100 individuals. Chen reflects on his experience at the once-fledgling startup:
“There was definitely a very magical vibe from the very beginning. OpenAI collected a group of extremely ambitious, talented AI scientists and researchers to think big and push the boundaries. OpenAI stood out as one of the few organizations thinking about that big leap forward for general intelligence, and that’s really the mindset that has created this movement behind the team, and ignited a momentum that has been instrumental in propelling the AI revolution that we are witnessing today.”
Was Chen surprised by the breathtaking speed of ChatGPT adoption by the mainstream?
Chen alluded to two significant surprises that have resulted. 1) The first was the rate at which technology has progressed. 2) The second was the swift adoption of these advancements. Chen highlighted the basis of ChatGPT’s success: the foundation model that trains across a multitude of tasks to enable learning across diverse domains, thereby enhancing its capacity for generalization.
Chen noted, “It’s trained on the Internet of text plus textbooks plus Wikipedia, plus OpenAI’s own proprietary datasets. And it’s by training on all these diverse sets of data which really represents different language tasks. It’s the ability to learn across all of these using a foundation model approach that makes it so powerful.”
Chen expressed his surprise not only at its existence several years ago but also at the rapid pace at which it has advanced. While he anticipated such a trajectory, the speed of progression caught him off guard. Chen elaborated, “Part of the reason why ChatGPT had the ability to gain popularity and adoption so quickly was because it worked on a lot of the tasks that people asked it to do, even if those are things the AI was not specifically trained on.”
Chen contextualized this rapid progression by highlighting the shift from task-specific AI to general AI, facilitated by the foundation model. In the past, developing AI models for specific tasks necessitated a team of experts, task-specific data collection, and costly training processes. This barrier hindered mainstream AI adoption. The shift toward general AI, exemplified by ChatGPT’s foundation model, has dismantled these barriers and enabled widespread adoption across various applications.
Chen drew parallels between the trajectory of advancements in language domains and robotics. He cited that before ChatGPT’s emergence, there were existing applications of language AI, such as Google Translate and sentiment analysis. However, the introduction of a general AI for text, like ChatGPT, ushered in a new era of expanded possibilities and value. He envisions a similar trajectory for robotics, envisioning the transformation of the current robotics market into a substantially larger one once general AI for robotics emerges,
“What we’re really seeing now is like what I would consider the pre-ChatGPT moment for language AI. We actually expect to see a lot more explosive growth as we view this general AI for robotics in a way that is similar to ChatGPT.”
Leaving OpenAI to Redefine Robotics towards Adaptable Intelligence
When Chen had decided to leave OpenAI, the company’s focus at the time was research. Chen and his cofounders embraced a different perspective. They believed that to advance general AI technology, it was crucial to create tangible products and deploy systems within real production environments. This would enable them to gather valuable data and insights for refining their AI models. Eventually, OpenAI transitioned towards developing a product, a public API and user interface, while leveraging real-world data and user feedback to enhance the technology. By then, Chen was already thinking about redefining robotics towards adaptable intelligence.
“Robotics has long served as a foundational element in modern manufacturing, historically confined to executing repetitive movements based on programmed instructions,” Chen explains. In the earlier iterations, robots were confined to specific tasks, operating within a rigidly structured environment. The limitations were significant; any deviation from their set tasks or alterations in the environment could render them ineffective.
However, the landscape has evolved, giving rise to what Chen refers to as “AI Robotics” or modern robots. These robots integrate the precision and reliability of industrialized hardware with a crucial addition—a cognitive AI component. This AI “brain” — aptly named at Covariant as the Covariant Brain — equips robots with the capability to comprehend their surroundings, make informed decisions, and adapt their actions according to changing circumstances. This pivotal advancement is poised to unlock a multitude of potential use cases that were previously inaccessible.
In the current scenario, a vast array of tasks requires a level of adaptability that traditional robotics could not offer. Consider the simple act of picking up a phone—depending on its placement, the required motion varies. Such nuanced tasks demand a robot to possess cognitive abilities and the capability to perceive its environment dynamically. The amalgamation of AI and robotics addresses this requirement, enabling robots to perceive, understand, and react differently based on the situation at hand.
Chen envisions a future where AI acts as the catalyst for an explosion in robotics applications. “You want robots to have a brain to see the world and understand it and make a different movement every time,” he emphasizes. This paradigm shift hinges on the realization that a single foundational model, a general AI platform, can underpin robots across various locations and tasks, enabling them to navigate the world intelligently and autonomously. Unlike narrow AI and the search for patterns in a defined manner, developing a generalized AI means being able to handle anomalies within the environment.
Jordan Jacobs, managing partner of Toronto-based Radical Ventures and investor in Covariant, was impressed by the company’s tech’s ability to receive instructions and pick items out of a bin and prepare them for shipping. This is a “valuable task to automate”, however he alluded to the challenges: “Developing an AI system that can operate a robotic arm accurately, then identify things from a pile of jumbled goods–Upside down, sideways leaning, and then get it right–that’s extremely hard.”
Chen explained the intricate challenges posed by AI: “The heart of this challenge lies in the ever-present variability of scenarios that must be addressed.” This inherent diversity, as Chen outlined, prompted the genesis of Covariant Brain with a focus on warehouses and logistics—an environment characterized by constant flux. Warehouses, by nature, encounter a barrage of changes—ranging from the introduction of new products daily to alterations in packaging and potential damage to existing packaging. This dynamic environment has posed an array of novel problems that robots must grapple with. The sheer multitude of scenarios that demand adept handling requires the capacity to navigate each one with a pronounced level of autonomy. “High throughput and pinpoint accuracy,” Chen emphasized, are the dual pillars that underscore this undertaking, presenting potential formidable AI obstacles.
Chen explained the rationale behind the Covariant Brain’s foundation model’s efficacy, “Covering various industries—from pharmaceuticals and cosmetics to fashion apparel and groceries—enables our AI to garner a comprehensive grasp of the world.” This breadth of exposure empowers AI with a generalized understanding of diverse contexts. When confronted with new and unfamiliar situations, the AI leverages its rich reservoir of past experiences to discern parallels and patterns. “Even though this exact circumstance has not been seen before,” Chen elaborated, “the AI has encountered countless similar instances, equipping it to tackle the new scenario adeptly and with a high degree of proficiency.”
This is the core of technology: an adaptable AI that thrives in an environment characterized by unpredictability. Chen’s vision is to imbue robots with the capacity to conquer novel challenges with precision and autonomy. He clarified, “The ability to learn across diverse tasks gives our AI a generalized understanding of the world, enabling it to handle novel situations effectively.”
AI, the Risk to Humans and the Future of Jobs
The implication of the next generation of robots on the future of human work is a common concern shared by many. Chen asserts, “We at Covariant firmly believe that robotics represents an overwhelmingly positive force for our society and the human race.” This conviction is grounded in a pragmatic examination of the challenges faced by various industries. The adoption of AI Robotics is often driven by the difficulties companies encounter in sourcing individuals willing to engage in repetitive and physically demanding tasks within warehouse environments.
He presents a compelling argument highlighting the unsettling reality that jobs characterized by monotony and high injury rates frequently exhibit annual turnover rates.
“Warehouses, often full of moving machinery carrying heavy loads, are inherently one of the most dangerous places to work. If you’re a warehouse manager, failing to control these risks can lead to high employee turnover, decreased productivity, legal issues, and, at worst, injury or fatality.”
The U.S. Bureau of Labor Statistics (BLS) reports that the warehouse and storage industry employs over one million workers in more than 17,000 locations across the nation. 5% of all warehouse workers in the US have one on-the-job accident at least once a year. This is an industry with the second highest turnover rate of 40% in 2021. “More figures from the Liberty Mutual Research Institute for Safety estimate American companies lose $62 billion per year due to workplace injuries. Most of them are due to safety violations causing expenses from lost time, medical treatment and disability payments.”
“AI Robotics,” Chen states, “emerges as a transformative solution, enabling workers to shed the burden of these monotonous and injury-prone tasks and transition to roles that are more stimulating and fulfilling.” He envisions the new role for human workers involves overseeing teams of robots and optimizing their operations. He argues, “AI Robotics are about enhancing human capabilities, providing more strategic roles rather than replacing human involvement entirely.”
What is Chen’s view on AI’s risk to humans? He maintains that the controlled environment of AI Robotics, particularly within warehouse settings, offers a natural boundary that minimizes the risks associated with arbitrary and unrestricted AI capabilities. Industrial robot arms operating within predefined parameters and safety enclosures provide a secure framework for AI deployment, compared to the adaptability of AI systems with broader online access in work applications, social media, and the web.
Chen further emphasizes that AI Robotics carries a lower risk of technology misuse compared to generative AI technologies. Addressing the concerns of misinformation and fake content generation as critical challenges associated with certain AI applications, AI Robotics, in contrast, are confined within a controlled operational scope, mitigating the potential for misinformation dissemination and aligns with a mission to promote productive and secure applications. He points out, “Our robots are designed to deliver labor savings and reliability improvements, presenting a strong business case for adopting this technology.”
Are there quantifiable savings and performance results from the implementation of the Covariant Brain in client environments?
Chen affirms, “Certainly, that holds true for all our clients. When considering warehouse operations, their cost structure is subject to intense scrutiny.” Chen emphasizes that warehouses, traditionally regarded as cost-centers, have now evolved to become differentiators in terms of capabilities. He validates, “We commonly see double digit labor savings and improved productivity with our customers. Whether that’s high double digit or low double digit depends on the customer.”
Nurturing an Innovation-Driven Culture
How does a company thrive in an environment of chaos–beset by perpetual change and uncertainty? Moreover, does it take a unique individual to work within this setting characterized by constant transformation? For Chen, his overarching philosophy means fostering a culture of continuous innovation, underlining that Covariant operates in a domain where simply making incremental improvements is insufficient emphasizing, “Our focus is inventing new solutions and doing so consistently.”
Through the hiring process, the company actively seeks out individuals who challenge prevailing norms and relish the opportunity to contribute to core innovations. He emphasizes the importance of crafting a deliberate culture that not only fosters skill acquisition but also promotes an eagerness to embrace the unknown. “A culture of fearless exploration is crucial,” as Chen pointed out, “because innovation inherently involves risks and defying the boundaries of conventional possibilities.”
The team and their shared values lie at the heart of Covariant’s ability to maintain a trajectory of continuous innovation. Chen believes they have a penchant for unconventional thinking and audacious problem-solving by prioritizing learning and exploration. This culture serves as the bedrock of change that empowers Covariant’s team to confront evolving challenges.
Chen envisions 2023 as a pivotal year, marking the imminent “ChatGPT moment” for robotics. Covariant Brain’s AI-powered robots are poised for extensive scaling across diverse industries. The upcoming year is predicted to usher in rapid growth and transformative impact, serving as a testament to the culmination of a journey initiated—an era where AI Robotics come into their own.
Perhaps, this is a pragmatic revolution in progress. The fusion of AI and robotics is set to drive profound changes, reshaping industries and redefining human interaction with technology. With AI-equipped robots expected to be commonplace, the potential for adaptive learning and efficiency becomes clear. The rise of the AI robot is unlike what we’ve only witnessed in movies, but rather, as per Chen, the symbiotic relationship between robotics and human endeavours, where technology serves as a catalyst for enhancing work experiences while minimizing risks to the work environment.