Seafarers carry the risks for maritime’s AI transition without having any control over the impact on ship-board operations or legal liability, writes Officer of the Watch Vladyslav Sidukhov.
It’s 03:00 AM, raining heavily, and the Malacca Strait is packed with traffic. The radar is cluttered with rain squalls. Suddenly, the multi-million-dollar AI (which is likely to have been hyped up when it was being sold) ‘collision avoidance’ system on the bridge decides to freeze or worse, it suggests a hard-to-starboard turn straight into the path of an un-lit fishing boat.
As the Officer of the Watch, I have exactly 10 seconds to recognise the machine’s error, disengage the autopilot, take the wheel, and save the ship.
Intuition and timing are what matters. If I succeed, nobody claps; it’s just another Tuesday night. But if I fail, the algorithm doesn’t go to jail. The software developer in California doesn’t lose their license… I do. Or my captain does.
I can write a whole guide or assessment on ‘How to lose your job while on a ship’.
The boardroom dream vs. saltwater reality
Whenever I read the news from the maritime technology sector, I feel like the people writing it work in a completely different universe. In the shiny boardrooms of London and Singapore, artificial intelligence is presented as a magic wand. The corporate pitch is beautiful: AI will optimise our routes, save fuel, eradicate human error, and eventually pave the way for fully autonomous, crewless ships.
But down here on the deck – in the real world of saltwater, heavy vibrations, and extreme fatigue – the picture is vastly different.
We are currently stuck in a very dangerous ‘grey zone’. The industry is forcing ‘semi-autonomous’ technology onto our bridges. These systems are smart enough to interfere with navigation, but they are nowhere near reliable enough to be trusted blindly. When a 10-meter wave smashes into the bow and blinds the external sensors with sea spray, the AI doesn’t step up to save the day. It throws a red ‘SYSTEM ERROR’ on the screen and leaves the mess to the exhausted human officer.
Yes I know that the example I gave was a worst-case situation that does not happen every single day, but everyone forgets that AI is first and foremost a tool and assistant that needs to be validated based on one’s own knowledge and life experience. Personally, I haven’t yet seen AI being implemented in the shipping industry on a large-scale, widespread basis, except perhaps through personal use, which has helped me solve many problems that benefitted from access to a neural network.
The ultimate legal shield
What really keeps navigators awake at night isn’t the technology itself. We are seafarers; we have adapted to everything from sextants to GPS. What terrifies us is the hypocrisy of the legal system. There is a sacred maritime law called the ‘Master’s Overriding Authority’, which unfortunately is increasingly losing its legal value. It means the Captain has the final word and the ultimate legal responsibility for the ship, the crew, and the environment.
But today, shipping companies are using this centuries-old law as a modern-day legal shield for their untested software (and simply because they don’t give a f*ck).
Let’s be brutally honest: companies are cutting crew sizes (down to minimum safe manning requirements) because they claim ‘the AI helps reduce workload’. But in reality, we are now doing the jobs of three people while also acting as full-time IT support for buggy algorithms.
And the moment that algorithm makes a fatal miscalculation, the shore management will point to the Master’s authority and say, “Why didn’t the human stop it?”
But how are we supposed to manage this if we have no overriding authority?
If they really want to outsource the calculations to the AI, then why not ask it to manage all situations, including the ones where the vessel is in danger? Suddenly the smart software is not smart enough!
You cannot code survival instinct
The fact is that you can program an AI to recognise a buoy, but you cannot code ‘seamanship’. You can’t teach a microchip to feel the slight change in the main engine’s vibration that tells you something is wrong before the alarms even sound. Machines don’t have gut feelings and cannot be astrologers. They don’t have a survival instinct. ABSOLUTELY f*cking not!
Seafarers aren’t afraid of the future, but we also don’t want to become slaves stuck in a critical period of AI’s maximum influence on everything, including the level of cortisol in the blood, the presence of ideas in the head, and so on.
We are just tired of being the scapegoats for machines that don’t know how to swim. If I had to simplify how we should think about it, if we are taking on the legal responsibility, then we need to be the ones in control. In equation form, skilled professional = human + tool. There is a lot that we can achieve with the right tools including AI, but humans and seafarers must be able to control these tools rather than being controlled by them.
The transition to autonomous shipping is happening, and crew members are not trying to stand in the way of progress. But before the industry spends another billion dollars on green-lit screens and predictive software, they need to fix the liability crisis. If an algorithm is given the power to steer a 200,000-ton floating steel beast, the developers, the manufacturers, and the shipowners must carry the legal weight of its actions.
Seafarers are highly trained professionals, we are NOT legal shock absorbers for Silicon Valley.
———–
Words by Vladyslav Sidukhov. Photo by Tom Donders on Unsplash
Click here to sign up for our free monthly ‘De-Mystifying Digitalisation and AI in Maritime’ newsletter and get the latest cross-sector expert insights, news, interviews, videos and case studies straight to your inbox.