Iceberg Mover

Moving icebergs is real work. The best-documented case is off Newfoundland, where icebergs drift toward the offshore oil platforms of the Grand Banks. The job has two parts. At sea, the crews of offshore supply ships lasso an iceberg with a floating rope or net and tow it onto a safer course, or push small ones away with a water cannon. A veteran captain told Legion Magazine that he sizes up each berg's shape, the sea and the weather, and that "each situation is unique." On shore, ice-management teams track icebergs with satellites, radar, aircraft and drift models, and decide which ones need moving. That article is from 2007; it describes how the work is done, not how many people do it today.

AI reaches the second part far more than the first. A 2023 study in the journal The Cryosphere trained a neural network to outline giant Antarctic icebergs in satellite radar images, cutting the time per outline "from several minutes to less than 0.01 s." It still missed parts of the very largest bergs. Tracking and forecasting work like this is likely to need fewer people over time. Towing a 200,000-ton iceberg in rough seas is hands-on seamanship, and nothing in the evidence shows AI doing it.

The government has no category for iceberg managers. The closest, captains, mates and pilots of water vessels, is projected to grow 4% from 2025 to 2035, about average. A 2025 Microsoft Research study of AI chatbot use ranked that job 433rd of 785 for how much of the work AI can help with, in the lower half.

We lower the score from 3 to 2. The earlier 3 rested on claims we could not source; the evidence we have shows AI helping with the monitoring while the vessel work stays human. What would change the score: automatic detection and drift forecasting good enough that offshore operators cut their ice-management teams would push it up. A wind-down of offshore oil on the Grand Banks would shrink this job for reasons that have nothing to do with AI.