Team
I think ACRs and pre-dreads are cool. So I asked ChatGPT to pose some hypotheticals for the Med circa 1900 (which would allow French, Italian, Russian, A-H and GB ships easily), using our hosts models. It came up with:
Option 1: RN ACRs vs. Russian Battleships & Escorts
Royal Navy: 3 ACRs (Cressy, Aboukir, Bacchante) + 1 Canopus-class BB
Russia: Slava, Tsesarevich, + 2 armored cruisers (Bayan, Pallada)
Tactical Angle: The RN must use speed and maneuver to avoid getting hammered by the Russian BBs while taking out the cruisers.
Option 2: RN Pre-Dread & ACRs vs. Russian Pre-Dreads & Cruisers
Royal Navy: 1 Formidable-class BB + 2 ACRs (Hogue, Sutlej)
Russia: Slava, 1 armored cruiser (Bayan), 1 protected cruiser (Askold)
Tactical Angle: More balanced in firepower; the RN BB can duel Slava while the ACRs try to outmanoeuvre and isolate the lighter Russian ships.
This may lead somewhere, though not all the ships it listed are available... AI is by no means perfect.
Mike
Chat GPT for hypothetical ~1900 naval conflicts
Moderators: dnichols, GHQ, Mk 1
-
Mickel
- E5
- Posts: 321
- Joined: Sun Jan 23, 2005 2:00 pm
- Location: Adelaide, Australia
-
redleg
- E5
- Posts: 3809
- Joined: Fri Dec 17, 2004 11:02 am
- Location: Riverside, CA
Re: Chat GPT for hypothetical ~1900 naval conflicts
That's pretty cool! I never thought of using AI that way. I wonder if you could then go on to have it play your opponent in one of those engagements?
-
Mickel
- E5
- Posts: 321
- Joined: Sun Jan 23, 2005 2:00 pm
- Location: Adelaide, Australia
Re: Chat GPT for hypothetical ~1900 naval conflicts
I suspect you probably could. But where is the fun in that? 
As I explored further, I found it was useful for coming up with scenarios and could then justify why these were at least plausible in a geo-political sense at the time, eg what the potential flash-points were, why, and where they might occur (open sea or more littoral environments). What it wasn't good at was knowing what ships fit where. I had to correct it on what ships were in what classes and sometimes what type of ship one was (eg ACR or pre-dread). But it appeared to make notes as we went, so that should improve. It was better with the battleships.
I might wind up with small numbers of random ships of the period, which will make for something different. There appeared sufficient tension that just about anyone vs anyone could have happened under certain circumstances.
As I explored further, I found it was useful for coming up with scenarios and could then justify why these were at least plausible in a geo-political sense at the time, eg what the potential flash-points were, why, and where they might occur (open sea or more littoral environments). What it wasn't good at was knowing what ships fit where. I had to correct it on what ships were in what classes and sometimes what type of ship one was (eg ACR or pre-dread). But it appeared to make notes as we went, so that should improve. It was better with the battleships.
I might wind up with small numbers of random ships of the period, which will make for something different. There appeared sufficient tension that just about anyone vs anyone could have happened under certain circumstances.