RALEIGH – Last week I wrote about concerns in the public that artificial intelligence (AI) will eliminate jobs. In particular, this emerging technology is expected to disrupt white collar industries like law, medicine and engineering. AI will inevitably eliminate numerous jobs that previously required human creativity and advanced degrees.
If you’d like to catch up on this story, you can read last week’s piece here. But the long and the short of it is this.
Throughout history, technology advances have created far more jobs than they eliminate; something economist Robert H. Solow modeled and proved mathematically, earning him the Nobel Prize in Economic Sciences in 1987. The number one driver of societal economic growth is technology advancement, far exceeding increases in labor or capital.
Today, I’d like to focus on what kind of jobs and industries will be most positively impacted by AI.
First, let’s consider the infrastructure of AI. Artificial intelligence requires massive amounts of computing. As such, there is a need to rapidly scale the number of data centers around the world. This leads to construction jobs to build data facilities, factory production jobs to build servers and IT jobs to install and maintain the centers.
The millions and millions of new computers and components will further grow industries at the foundation of the electronics supply chain like the mining of minerals and precious metals, and transportation of materials around the world. We’ve seen proof of this kind of net-plus job creation many times before. The telephone created logging jobs for telephone poles. Steam power increased coal-mining and steel production.
We will continue to network these data centers together, growing fiber optic production and installation, which is very difficult to automate. We will build more radio towers and satellites to wirelessly connect the furthest reaches of the planet to sensors that create data to feed AI and to people who benefit. We will retrofit our buildings with 5G access points, IoT sensors and edge computers. Operational Technology (OT) will scale and become as recognized as Information Technology (IT) is today.
AI consumes massive amounts of energy. Earlier this month, I joined the 10th anniversary of the Virginia Academy of Science, Medicine and Engineering (VASEM), where the head of the Department of Energy for the State of Virginia discussed future energy needs across the state. Virginia is one of the top three data center states in the US. Today, these server farms consume ~20% of the energy in Virginia. The energy secretary explained that one training run for a large language model like chat GPT consumes as much energy as 127 households use in a year.
AI will increase demand for energy sector jobs, expanding wind and solar production and grid-scale battery storage facilities. But it will also create new job categories, not yet fully understood. The utility industry will completely transform as the grid changes from today’s star-like network of uni-directional power transmission to end users. The future grid will be an AI-supported mesh-like network of bi-directional power sharing between distributed production, storage and use. The future of electricity distribution will look less like shipping cargo along a supply chain and more like distributed packet data transmission on the internet. Designing and managing this requires a whole new category of jobs, where people will augment their intelligence alongside AI tools and controls.
There will be additional new industry sectors that I predict will emerge. One way to think about this is to consider what new problems will need to be addressed, as AI scales. A few immediately come to mind.
- A great deal of AI happens in a “black box”. Explainability will be important, when AI generates something that has an unanticipated result. I expect an industry to emerge that will focus on forensic analysis of AI.
- Corporations will implement new C-suite jobs. We already see Chief Data Officers and Chief Information Security Officers. I anticipate new Chief Ethics Officers and Chief Privacy Officers to emerge, with teams focused on maintaining customer trust.
- Some uses of AI will drive inefficiency, even as they create efficiencies. Consider how many companies will use AI to generate outbound marketing campaigns that they can now scale to send to huge audiences that were hard to scale to previously. I’m already seeing a spike in the amount of inbound AI-generated email. In the same way the anti-virus industry emerged in the early days of the internet and spam-blocking industry soon followed, I expect a new industry of “AI mitigation tools” to emerge. Humans will be a part of tuning those algorithms.
- AI systems are already increasing the resolution at which we understand the health of physical infrastructure through things like drone-based vision inspection of power transmission lines and cameras mounted to vehicles to inspect roads and railways. The real-time maintenance of these systems can be greatly improved and I expect future technicians to no longer be manual repair-persons, but IT-enabled professionals in the field.
- We are likely to see nursing and child and elder care industries elevate significantly. People trained in the hands-on aspects of those industries will now have voice and vision-fed AI to augment their human skills with “on-call” AI assistants, greatly increasing the quality of care they can provide. These will be high paying jobs.
- There is a chance that education will transform from a model of 1 teacher per 30-40 students, instead to one where teachers work much more closely 1-on-1 with students who best learn through human engagement. At the same time, other students may be able to move forward more independently, with AI coaches and guides providing personalized learning. New kinds of coaches, mentors and subject matter expert jobs could replace our more generalized teaching positions of today – but in far greater numbers.
In the immediate future, I am not too concerned that AI will replace so many jobs that we don’t have enough work to keep people engaged. There will be new industries that rise up. The bigger challenge will be how we manage the change. People can be resistant to change, and re-training to a new career is intimidating.
But there is always a chance that this particular technology advancement is different. Maybe since so much of AI is machines working with machines, that this time we won’t see the number of new jobs created exceed the number of old jobs lost. What should we be doing now, in preparation for that possibility, even if it is remote? I’ll jump into that topic next week. Thanks for reading!