How do we learn to work with intelligent machines? | Matt Beane

TL;DR

This content explores how the current approach to handling AI in various industries is blocking on-the-job learning and suggests a need for a new approach.

Transcript

It’s 6:30 in the morning, and Kristen is wheeling her prostate patient into the OR. She's a resident, a surgeon in training. It’s her job to learn. Today, she’s really hoping to do some of the nerve-sparing, extremely delicate dissection that can preserve erectile function. That'll be up to the attending surgeon, though, but he's not there yet. She... Read More

Key Insights

• 🏥 Surgeons and other professionals traditionally learn their skills through a process of apprenticeship, mentorship, and on-the-job training, known as "see one, do one, teach one".
• 💡 The rise of AI and automation in various industries, including surgery, has disrupted this learning process, as AI systems are often designed to take over tasks and leave little room for hands-on learning.
• 🌐 This trend is not limited to surgery; it is happening across different industries, such as start-ups, policing, investment banking, and online education, where learners are being peeled away from expert work and struggle to gain practical experience.
• 💻 McKinsey estimates that between half a billion and a billion workers will need to adapt to AI in their daily work by 2030, relying on on-the-job learning to develop the necessary skills.
• 🔒 The current approach to AI is blocking learning on the job, hindering workers from acquiring new skills and adapting to AI technology effectively.
• 📚 Some individuals, labeled "star pupils," have found ways to learn in the face of AI's disruptions by breaking and bending rules, taking risks, and seeking expert support to tackle challenging tasks.
• 📍 To overcome this obstacle, a combination of struggle and expert support needs to be integrated into AI implementations, mirroring the success of those who engage in "shadow learning" by combining hands-on experience and guidance from experts.
• 🌍 Advancements in technology, such as the internet and cloud computing, provide opportunities to scale these learning approaches beyond the limitations of physical proximity, leveraging AI to coach learners and connect them with experts in smart ways.

Questions & Answers

Q: What is the main path to skill around the globe for thousands of years?

The main path to skill around the globe for thousands of years has been through apprenticeship, coaching, mentorship, and on-the-job training.

Q: How is learning being blocked in the field of surgery when it comes to AI?

In the field of surgery, learning is being blocked when it comes to AI by using robotic technology that allows attending surgeons to perform procedures themselves, limiting the opportunities for residents and trainees to actively participate and learn through hands-on experience.

Q: What is the potential impact of AI on the workforce?

According to McKinsey, between half a billion and a billion people will need to adapt to AI in their daily work by 2030. This highlights the importance of on-the-job learning and the need for it to be preserved and supported in order for workers to effectively transition and develop the necessary skills to work with AI.

Q: How can organizations ensure that on-the-job learning is preserved in the age of AI?

Organizations can ensure that on-the-job learning is preserved in the age of AI by incorporating a combination of struggle and expert support into AI implementations. This means allowing individuals to tackle challenging problems near the edge of their capacity while having access to expert guidance and assistance to enhance their learning experience. Additionally, leveraging tools like the internet, cloud, and AI coaching systems can further support on-the-job learning efforts.

Summary & Key Takeaways

• Surgeons in training, like Kristen, are struggling to learn and gain experience in the presence of AI technology, such as robotic surgery.

• Organizations in various industries are prioritizing productivity over learning on the job, which is hindering the development of necessary skills for workers.

• It is crucial to find ways to incorporate AI technology in a way that enables learning and growth, by allowing learners to struggle and providing expert support and guidance.