It's been done many times before, quite successfully. The Napier Deltic (3 crankshafts!) and Junkers Jumo are two examples. The Commer Knocker is another (which had only one crankshaft I believe). Two crankshafts makes the engine slightly taller, but not egregiously so.
I'd be curious to know more about the heat problem you've mentioned. Naively, there should be much less waste heat to dispose of due to the fact that the same cylinder volume is doing twice the work.
1. Stop requiring computers/phones for everything. Your 91 year old grandma isn't going to make her way through your super cool very intuitive 2FA magic link email confirmation system, and I don't WANT to make my way through your super cool very intuitive 2FA magic link email confirmation system.
2. teach the people who need to use computers, how to use them.
I never said anything about 2FA magic links? We can do much, much better via things like FaceID integrated passkeys, and probably further steps from there.
> Stop requiring computers/phones for everything.
Ah yes, that sounds straight forward. Let us know when you’ve deployed that to prod.
Not claiming anything to be false, just a reminder that you should question ones opinion a bit more and not claim they "know what they are talking about" because they worked with Fei-Fei Li. You are outsourcing your thinking to someone else which is lazy and a good way of getting conned.
I was very surprised to find the opposite yesterday. I was asking ChatGPT about firearms and it hit a safeguard ~”I cannot give gun purchasing advice” so I switched to Gemini, and it happily answered the exact copy/paste question
Historically it was the opposite; OpenAI was yolo and Gemini overly cautious to the point of severely limiting utility
If it’s anything like the electric ferries that cross the Öresund beween Helsingborg and Helsingør, they grab charge while they’re unloading and loading at each terminal:
Each trip consumes approximately 1,175 kWh, which is nearly the same amount a residential home consumes in a month. In each port is a tower with a robot arm that connects the charging cable automatically every time the ship comes to the dock. The system charges 10.5 kV, 600Amp and 10.5MW. The batteries have a total capacity of 4,160 kWh, which means that we always have a surplus of electricity if for some reason we cannot load during a stop or if the transit takes more time than usual.
In Helsingör the ferries charge for approx. 6 minutes and in Helsingborg the ferries charge for approx. 9 minutes. This is enough to suffice for the journey across the strait.[1]
Side note: you can also charge your car on board from the boat’s batteries.
The Cruise Ship Terminal in San Francisco has 12 mW. Apparently it's uncommon in that it's wired with enough power available so the cruise ships don't have to run their on board generators while docked in port here. It's a major pollution thing.
> The ship... will travel between the ports of Buenos Aires, Argentina, and Colonia del Sacramento, Uruguay. The two cities are 60 kilometers apart, a distance it is expected to travel in 90 minutes.
> Direct-current charging stations will be installed at each port... A full charge is expected to take just 40 minutes.
Something people overlook with these things is that you don't actually need to fully charge batteries because they won't be completely empty and probably a 70-80% charge is more than enough for a single crossing with a healthy safety margin. Also charging speeds are non linear. Charging speeds typically drop when the battery gets closer being full. Charging from 80% to 100% is a lot slower than charging from 20% to 80%. And depending on the battery chemistry, completely discharging or charging them to the max isn't necessarily great for battery longevity.
Another point with battery powered ships is that the rate at which they discharge is speed dependent and that's a non linear relationship because the drag increases quadratic with speed. So, if you are at 30%, you can still make it across. Just not at the full speed. This is less about range anxiety than it is about just being able to stick to schedules. If the ship did not charge enough it would have to go slower. But it would still get there. This ship is designed to go quite fast which means it would have a lot of wiggle room. So they might make it across at full speed even at maybe a 60% charge. The risk is that they'd run low and might have to slow down a bit. It would get there but with a delay if that happens. And then it would have to sit there a bit longer recharging leading to more delays.
The trick is optimizing the amount of batteries to minimize turnover and delays; not around being able to charge them from 0 to 100%. The sweet spot is probably around the 20-80% mark, meaning you'd want to be able do a crossing at full speed using about 50-60% of the battery capacity. The rest is just there as safety margin to avoid delays. If you burn into that, you need to charge a bit more. With 40-50 minutes turnover, there's plenty of time to do that typically.
Indeed, that's why I say "keep charge", i.e. be in a steady state such as always leaving at 80% charge. Not charging from zero, and not necessarily charging to 100%.
People who charge electric vehicles at home emphasise that you plug it in as a matter of routine every night (ABC: Always Be Charging) and since it's software-controlled, you can e.g. tell it to charge up to 80%, and figure out the most cost-effective way to do that by 8am.
The ABC of such a ship, is that it would be plugged whenever it is docked, during the turnarounds. And there is enough time in that turnaround to keep charge. It likely also has some downtime at night as well, but that matters less in this case.
There’s absolutely been a lot of focus on LLMs, but they simply work very well at a lot of things.
That said, Carbon (C++ successor) is an active experimental (open source) project. Fuchsia (operating system, also open) is shipping to consumer products today. Non-LLM AI research capabilities were delivered at a level I’m not sure is matched by any other frontier lab? Hardware (TPUs, opentitan, etc). Beam is mind-blowing and IMO such a sleeper that I can’t wait for people to try.
So whilst LLMs certainly take the limelight, Google is still working on new languages, operating systems, ground-up silicon etc. few (if any?) companies are doing that.
> Sorry, but AI still seems to be trash at anything moderately more complex than baby level tasks.
How familiar are you with the concept of the jagged frontier? That is, AI does indeed fail at things we might expect a third grader to be capable of. However, it is also absolutely exceptional at a lot of things. The trick is A) knowing which is which and B) being able to update yourself when new capabilities are unlocked
So yeah, it’s unsurprising you found a use case it couldn’t trivially do. But being able to one-shot quite complicated applications that may have taken a day to get right previously is an astonishingly useful thing, no?
reply