Too many years ago, we bought and renovated an old state house. As we painstakingly repaired the broken plaster, I learned about ‘lathe and plaster’ and the lost professions of plasterers wall-paperers.
As the house was built, builders covered the internal walls with lathe, thin strips of wood, for the plastering team to come and manually plaster. Which, in turn, the wallpapering team covered with lining paper and wallpaper. How many full-time plasterers and wall-paperers do you know today?
Two days ago I had a coffee with the founder of a US startup that’s using AI to do customer phone sales. It’s great technology. I trialled the AI and was impressed by its sophisticated ability to converse with me and guide our conversation.
Another startup, Sierra.AI, founded by serial entrepreneur and former CEO of Salesforce, Bret Taylor, is using AI to automate customer service. Sierra already has some big customers such as Weight Watchers.
Yesterday I watched a YouTube video of the University of Maine printing a house with bio-based materials. No builders, plasterers or wall-paperers in sight. Recyclable.
These products are taking over jobs done by people. Builders. Customer service staff. Sales people.
The common response is that firms will get more output from staff rather than use less humans to do the same output. For example, customer service is an expensive service to provide to customers without any new revenue from providing it. So the AI can do something extra that firms wouldn’t otherwise do.
Hmmmm. Yeah, nah. Jobs will be lost. As a firm I may be able to do more customer service, but I’ll inevitably reduce my spending on human customer service personnel.
As a society, we need to start thinking now about the jobs we need to create to replace those lost. We must think about the regulatory framework needed in this environment, and the training pipeline for a generation growing up with AI.
It’s not the role of governments to plan future markets, especially labour markets; something which they’ve been historically bad at. However, a hands-off approach to market regulation can lead to exploitation of workers in the ‘gig’ economy. A live issue globally.
What about education? AI is moving into data analysis, lesson planning, resource development, marking, teaching (turn the online sales rep above into an online teacher) and student feedback. There’s no point debating whether AI should do that work. It’s happening.
The more important questions are what’s the role of those tasks in a teacher’s ongoing development and how will automating them affect this?
Good data analysis helps a teacher process what the data means and how they’ll respond. Lesson planning gets teachers to think about their use of class time, personalising activities for their students, and building in time for individual interactions.
Most people agree that we all want teachers in future classrooms, regardless of AI. But what will be their professional autonomy? Will future teachers deliver scripted lessons or script their own lessons to deliver?
AI can and will reduce the mundanity of teaching. It could also increase it. We must start thinking now about how we’ll retain professional accountability, autonomy, creativity and relationships in an AI age.
And we must design solutions and frameworks accordingly.
Comments