AI in Education: The Future of Work Will Shape The Future of Education
Inaction might render traditional educators irrelevant and turn beautiful campuses into the abandoned shopping malls of the future
If we designed our education system from scratch today, it would look radically different from the antiquated, expensive, slow, and often stuffy system we have. As an AI futurist, I spend time researching and thinking about the future of work, education, and society. The question we must ask ourselves is not only what will happen when we put AI in education but what needs to change in our education system to prepare for AI and society in the future.
Our formal education system was designed during the Industrial Revolution to help farmers transition into factory and office jobs. The curriculum was designed to teach them basic math, communication, and science skills they needed to be productive workers—as engineers, machine operators, construction workers, and so on. Vocational training prepared people to work in the trades as carpenters, plumbers, and, later, electricians. Incidentally, schools close for long periods over the summer because 18th-century students needed to do farm work and handle harvest during that time. The past always has echoes.
First-world economies in the early 21st century focus on manufacturing, retail, and service industries, and society’s educational needs have shifted. The curriculum has adapted to meet these needs, and some students now go on to higher education to gain the skills needed to participate in higher-paid jobs in marketing, finance, and high-tech. Education is still incomplete, however. Many students emerge from their education without vital life skills such as how to manage personal finances, navigate corporate life, or deliver a persuasive presentation.
While the curriculum has adapted, pedagogy remains essentially the same: a room full of students sat behind desks, facing a teacher who shares pearls of wisdom lecture-style and sets assignments as homework. Most schools have high student-to-teacher ratios, and the teaching pace is geared to the average student, which leaves some students struggling to keep up and others bored out of their minds. The whole process occurs at a sedate pace. A graduate degree takes three years to complete in most European universities and four years or more in the United States. Students with aptitude could complete degrees far more quickly and significantly so if aided by AI tutors. But we will get to that in a minute.
Buckle Up: Massive Change is Coming
Society is about to experience a considerable transition. Artificial intelligence will bring at least as much change to our lives as the original Industrial Revolution did to our ancestors. In the 18th and 19th centuries, we built an educational system to meet the needs of the Industrial Age. Now, we must build a new system to prepare society for the Intelligence Age.
Rapid innovation will quickly drive the cost of powerful intelligence down towards zero. By the end of the decade, humanity may have created artificial general intelligence, able to perform almost any task we set it. Work will transform forever. Many old jobs will be destroyed, and new job types will be created. It will be a jarring time. People will need support to prepare themselves for an alien world of work where machines are coworkers and perhaps also their bosses. We need an educational response of the same scale as the one summoned for the Industrial Revolution. This new educational system won’t be built overnight. But we need to start imagining and building it right now.
Work in 2030 will look quite different than it does today. We may even witness the end of labor within our lifetimes. The Chief of Staff to the CEO of Anthropic, a frontier AI company, believes that with additional rapid improvements in AI, she may be experiencing her last five years of work. It’s unlikely we will transition to a post-labor world that quickly, but it’s feasible to consider that most jobs people do today will have vanished within a generation.
AI and Automation vs Augmentation
You can summarize the future of work in two words: Automation and augmentation. AI and robotics will fully automate some roles while others will be augmented. Collaborative AI agents and cobots (collaborative robots) will boost people’s productivity, extend their abilities, and improve the quality of their output. Machines will transition from being tools to coworkers. That process is already underway.
AI Assistants Will Transform Knowledge Work
Today’s chatbots will quickly evolve into full-blown AI assistants, agents that can handle complex tasks and use tools to get things done on our behalf. Specialist agents will support workers in various domains, including clinicians, lawyers, HR specialists, financiers, engineers, marketers, salespeople, and filmmakers. Agents will handle lower-value work and elevate human performance on higher-value tasks. Assistants will help us conduct research, create content, analyze data, and make decisions. They will boost our senses, aid our intuition, improve communication, and increase productivity.
AI capabilities are growing exponentially. In a phenomenon known as ‘emergence,’ the abilities of large language models like GPT-4 tend to increase as the models' size grows. Leading AI models have grown by an order of magnitude in size every two years. GPT-4 is roughly ten times the size of GPT-3. GPT-5 (or whatever it will be called), likely coming in late 2024 or early 2025, will be at least ten times the size of GPT-4. In this industry, size really does matter. Dario Amodei, CEO of Anthropic, believes that models will continue to yield scaling benefits and become at least 1000 times larger than they are today, perhaps 100,000 times larger. This may result in AI that’s powerful enough not only to support people in their work but eventually to displace them from it entirely.
Robots Will Transform Physical Work
The pace of innovation in robotics is breathtaking. Humanoid robots have improved in leaps and bounds over the last 18-24 months. Advanced sensors, lightweight electric motors, robotics, and advanced AI have allowed leading companies Figure, Boston Dynamics, Agility Robotics, Tesla, Sanctuary, and others to develop robots that existed only in science fiction just a decade ago. These robots observe human behavior and then practice a task to learn how to do it. Gone are the days when you had to program the robot step-by-step to complete a simple task. The integration of powerful multimodal AI from OpenAI lets the Figure 01 robot see and perceive the world, understand verbal instructions, and even explain its actions. If you’ve not seen this demo yet, it will blow your mind. Remember, these robots are the worst they’re ever going to be, and over the coming few years, we should expect they will gain additional dexterity, mobility, and intelligence that make them highly attractive for manufacturing, warehousing, and repetitive manual tasks. The Figure 01 robot is on trial inside BMW’s manufacturing plant in South Carolina, and Agility’s Digit robot is at work inside Amazon warehouses. The world of physical work will look very different by the end of the decade, and the future global market for humanoid robots is estimated at 10 billion units.
The Future of Work Must Shape the Future of Education
As augmentation takes hold, students must learn how to thrive in a world of machines. Office workers of the 1990s learned to use computers, word processing, spreadsheets, and other software. Workers of the 2020s must learn how to get optimal results from artificial intelligence—how to use new AI tools and, soon, how to achieve optimal results working with AI assistants. These are vital skills for any worker. Schools should incorporate relevant AI training into the curriculum from an early age.
Sam Altman, CEO of OpenAI, speaking recently on how education needs to change to prepare students for the future, said, “People should not only be allowed but required to use [AI] tools. You need to understand math but also be proficient in how to use a calculator, too. [AI] is going to be an important part of doing valuable work in the future.”
As automation takes hold, most adults will need to upskill or reskill to maintain relevance in the workplace and prepare them for the subsequent phases of their careers. The education sector must reorient itself around lifelong learning, not simply preparing younglings for a monolithic career in a single job or industry, as they have previously done.
The cost of a quality education has become crippling for many, saddling students with a lifetime of debt. We must cut the cost of high-quality education by 90% and scale it to address the needs of everyone who needs access to it, not just the wealthy or those willing to take on massive debt in the hope of gaining access to a more lucrative career.
Artificial intelligence is a great amplifier. In the future, we may see billion-dollar companies run by a single person supported by a team of highly intelligent machine assistants and robots. Educators can amplify their impact, too. Students will always benefit from a deep connection with a human tutor, mentor, and guide. But AI tutors can and must play an important role that compliments human tuition and speeds learning with adaptive teaching tailored to the unique learning style of each student.
If AI could liberate human faculty to spend more quality time with more students (either 1-on-1 or in small tutorial groups) and relegate much of the 100-to-1 lecture-style tuition to 1-on-1 AI, it might be possible to amplify their productivity by 300 to 500%. If faculty and academic administrators embrace AI-powered tools to support everything they do, they might boost their productivity by another 200-300%. AIs can already mark papers, build complex class schedules, answer routine student questions, create course materials, brainstorm curriculum, create quizzes, manage finances, and act as teaching assistants. Within a few years, students will experience AI-generated virtual field trips that feel like the Star Trek Holodeck. AIs will assess individual student attainment in real time to tailor the pace and focus of their education. Faculty will remain fully appraised of each student through comprehensive AI-generated dashboards so they can determine where they need to focus their time.
The future of education can and should be very different than today. Our existing education system has served us well for the last 250 years, but there are better approaches to meet the needs of the future. Significant disruption is coming.
Academia may keep its head firmly thrust into the sand for as long as possible. Many in education will throw up hastily constructed reasons to ignore this warning, characterizing the author as naïve or lacking understanding of what it takes to provide a top education. This is denial at best and arrogance at worst. Schools and universities are incumbents in a sleepy market that hasn’t changed much in centuries, something academia prefers to label as ‘tradition.’ If they ignore the rapid pace of technological change, the spiraling costs of education, and the shifting needs of the job market, educators will seal their fate. When new players enter the market with compelling alternatives to expensive, slow, traditional education, it will be too late. Their goose will be cooked, and they will deserve to go out of business.
Perhaps the conventional education sector is incapable of change, too steeped in convention, high on their self-importance, and blind to the freight train of change bearing down upon them. But this is certainly not the case everywhere in education. Academics are a curious bunch, and I’ve spoken with several colleges and universities that readily recognize the need to adapt and embrace fundamental change to remain relevant and survive. Their forward-looking, savvy leadership recognizes the exciting opportunity ahead and the potential to scale their impact and fulfill their mission in exciting new ways.
The question for academics is simple: will you be an ostrich or a trendsetter?
Steve Brown is a leading AI expert with decades of experience in artificial intelligence and high tech. He grew up in an academic family, and his father is Professor Michael Brown CBE, former Vice Chancellor of Liverpool John Moore’s University.