I would not, no way no how. The pilot has skin in the game, he dies a 1/4 second before everyone else if he flies us all into a cliff. Some geek in a windowless room programming planes, well he won’t WANT to wreck them but it isn’t like it would hurt him to do so.
BTW - I just delivered an airplane across the country. It could do a lot by itself, but it was in no way close to being able to be just let loose with no humans. All that fancy automation has the occasional issue 
Not necessarily. Us "old guys’ always bitch about the “Children of the Magenta Line”, but I still teach all the old school stuff and some of us still like to actually know how to fly vs. giving the auto pilot dumb looks while it does random crap.
Keep in mind airplanes fly from between say half an hour to most of one day between mechanics being able to access the engines if needed.
If you want a ship to run like a modern turbine airplane, you would have the following on the bridge.
A switch with off-start-run.
A throttle and gearshift.
Another engine.
Most malfunctions would be taken care of by some automatic process or by securing the engine and using the other one.
Is this doable?
Another autonomous boat make a test run successfully.
OK, so it is not cross Atlantic, but running a narrow canal can be just as challenging, autonomouwise.
Although I haven’t flown many times since my retirement in 2003 at age 48, My rare trips since then were to New York to watch my son play football at USMMA, I drove most of the time and challenged the NJ turnpike and the ninja motorcycycle assholes. Imagine my sordid delight when someone opened their door at an inopportune time for the bikers on a certain tunnel during my transit. My Harley friends don’t do that shit. I do have preferences if I ever do fly again… A human on the bridge does give me some degree of comfort. Whether it be tugs,ships or airplanes.
the propulsion will have to change to make it work, no question.
are you teaching ppl or cpl, with cpl its the companies rules, they dont wan the crew to have skills and think like the old days
yes sure, but recent aircraft stats show the pilots crash the planes not the technology or failure so in the old days you were safer with pilots but not any longer, unless they are the grey haired old versions
You may be overthinking this. From say 1920 to now, the share of airplane incidents caused by pilots vs. caused by mechanical failure has been steadily increasing. This is not because pilots have gotten steadily worse at their jobs, it is because airplanes have gotten much better.
A pilot in say 1925 didn’t train for engine failures as something that had a very remote chance of happening, it WAS going to happen and probably way more than one time in a career. The old Jennys that were a staple of flight training and barnstorming back then had horrendously unreliable engines. Any still flying in 2021 pretty much fly around over the field they took off from, the only one I know of that flies cross country has a modern engine.
So now in 2021 for a commercial pilot flying cargo or passengers in modern turbine airplanes is not going to routinely be shutting down engines, they could easily go a whole career without doing so. The wings mostly stay on, the radios mostly work, the GPS mostly tells them where they are, and it takes some considerable doing to have no power on the A bus, the B bus, or the emergency bus.
There most certainly are problems with automation making pilots complacent and/or losing basic skills, but at least in most of the 1st World we are a long way from “They just type, if they have actually fly they’ll crash”. See the Korean Airlines pilots that hit the seawall landing visually at San Francisco and wrecked the plane, this would have been an utterly routine landing for USA based airlines.
BTW - I train my students to land at night with EVERYTHING failed. It isn’t easy, but it is certainly doable and you never know when you’ll need those skills…
yes and no
planes are getting better but stupid pilots are also get more stupid as the training skills are slowly dropping away.
Have a look at the accidents in the last 10 years, if the pilot wasnt there the plane would not of crashed in most cases, especially in Asia
In the “good old days” when men were men, women were men, and sheep were scared airplanes crashed a LOT more often than they do now.
Also if you dig a bit deeper, you have issues with not being able to solve issues caused by automation, not just lack of ability to fly at all.
One really has to dive deeper into systems now than ever before. The classic 6-Pack Steam Gauges on everything from 2-seat Cessnas to 747s back in the day were pretty easy to troubleshoot, everyone knew what hoses and wires went to what gauges and what problems they could have. Now there are a bunch of sensors, a bunch of wires, a big almost unknowable blob of computers in the middle, and a bunch of displays. Exactly how they interact with each other is not always obvious and some failure modes can produce some very bizzare outcomes.
On the flight I just did there were THREE different fuel quantity systems that ended up not interacting the way we thought they did. We figured it out, but back in the day this would not have been an issue.
Nice article from the WSJ to put some perspective on reality vs hype regarding autonomous technology
I have some background in the tech and I never had a clue how these self-driving cars were going to work. Hell, my car can lose lock on the car ahead if the curve is tight enough and nail the throttle because “the obstacle is gone” 
Thought it was just my rather new vehicle that did that. Works “most of the time”.
I’ve had the pleasure of working with some of the finest marine engineers in the business.
Nothing new, on the SAE blogs there are many who are posting that work in the industry and aircraft guys, they are all saying it will take years of testing before a test version should be allowed on the road.
They also know that as humans can anticipate and computers cant the public will need to get used to seeing people run down when they know a driver would have avoided the issue.
It will save many lives on freeway driving but might kill a few more on local streets where pedestrians are.
The first big hurdle is everything aware issue, it will only work when all the cars talk to each other.
Slocum, I concur. We should all be so lucky. First drink I bought in the airport going home was not for me.
Cars will have to talk to each other for any self-driving scheme to work. I am sure you have all experienced cruising along radar-locked to the car ahead of you when some traffic malfunction 3 or 4 or 10 cars ahead has everyone nailing the brakes. A human sees this and hits the brakes, the car does not see this and continues on full speed ahead until the car just ahead slows down. It is simple physics that if all the cars are on Otto, each one has to brake a little harder than the one before it until you finally get to the one that can’t stop in time.
yes everything aware will fix this issue as all the cars brake at the same time