Professions and Progression
People obsessed with Progression want to see numbers go up forever. People obsessed with Professions often want to feel purpose in whatever they make or do. These are the people likely to talk or complain the most. As Devs, the struggle is real. 'Casual' readers, worry not, you can just stop at the 'And Now, The Numbers' heading of this post, since this is all so we can reference 'things AI probably should and shouldn't do'. The examples in that section are just so people can equate stuff ingame to them.
Devs don't have time to create infinite engaging content, the best they can often do is randomize and connect things to the world and hope the player's imagination does the rest. The 'worst' they can do is pad it out with timesinks, because the obsessed Progression player won't complain as much, but normal players do. Another 'bad' option is to turn in products to NPCs for basically the same currency every time, because this turns one's Profession into a grey homogenous blob, especially without additional tweaks. This is all very hard.
Things Analytic AI can do to help
Analytic AI can sometimes do some basic number crunching in the spreadsheet to ensure that no particular profession is being overly rewarded compared to others., and this works in both 10K and UBI style games, though complex UBI games will have some issues if they don't commit fully to it, since the use of things by players or quests is a bit rough.
Example: In FF11 there is a quest that is triggered by having a particular bed in your Mog House, and requires a Shrimp Lure among other things. Neither the bed nor the Lure is super easy/nonspecialized to produce. This works because it is a 10K game and the Devs don't have to worry about every little piece of the Econ puzzle. With enough dedication, as I remember it, almost everyone could make that Bed, or that Lure (both certainly take meaningfully more dedication than 'just dabbling', but not clamped/restricted).
In a UBI game, the Analytic AI might miss this because the bed doesn't require more materials than other similar items but the player demand is a bit higher, so the 'reward' is higher (really, 'dependent on how many people see this and go for making them, making it a 10K style thing within a UBI game, which might sound good, but I assure it is not).
Things Generative AI can do to help
Mini Quest/Contract/Commission writing, obviously, also keeping things 'fresh' by tweaking the quest dialogues a bit on updates, allowing players to have new experiences akin to World Events and similar. For Ashes, Throne and Liberty, and FF11, these could all tie to various world states (FF11 has 'who controls which Regions this week', TL has 'which Dimensional Trial threat or event is active this week', Ashes has both Region control and Story Arcs as options).
Example: A pair of NPC Commissions that normally ask for a Rocking Chair and a Reading Lamp during 'whatever equivalent of Peace time' might change or swap around to ask for Hanging Lantern and Outdoor Chair of some kind during a Conflict period, precisely maintaining the overall balance of progression for the Carpenters and Lampmakers.
Things Devs Probably Have To Do Themselves
Niche Equalization mentioned earlier in the thread. Throne and Liberty has now completed their Niche Equalization on certain parts of Fishing (through another system that they overhauled, just think of Amitoi Expeditions as similar to FF11 Gardening I mentioned before, and now considerably moreso). The missing ArtP2 slots for gathering from Soil and Rocks, along with some related ArtC1 (technically) which somewhat equalize the 'professions' (assume they added 'House Construction/Repair Person' and it matches the approximate level of 'Lampmaker').
Analytic AI could handle things like 'how much exp should this give, how many materials should this cost, how much do we expect the players to want these items that have limited total use for any given player, how much do we think people will actually specialize in this based on time/progression telemetry?'. Generative AI could handle the 'writing of the Commissions for repairing NPC houses or installing windows so that the Econ segment keeps flowing without relying entirely on player Demand'.
But it's the Devs that had to trigger the initial Profession changes. It's the Devs that have to decide how much to include or exclude the Niche/Profession of 'Adventurer' or 'Dungeon Delver' from the process of "House Construction/Repair". The AI Agents can't 'see the world' as clearly as the World Design Team or 'see the intended Narrative' as well as the Narrative Design team or even 'see the player behaviour trends' in the same way the Econ Design team can.
And Now, The Numbers
I was going to just leave this part off, because the entire point of the prior half of this post is that it's different for every game. TL doesn't have 'specialized Bed-makers' on the same level that FF11 does. FF11 doesn't have Lampmakers to the level that TL does (speaking as technically one myself), blah blah. Hopefully speaking abstractly is still 'helpful'. To reduce some of my tendency to over-abstract I've asked some guildies to help me from their perspective so hopefully it won't end up being the usual.
Assuming the specific Equation of (Supply/Risk vs Demand/Reward), and yes, those slashes are 'divide by'…:
Fish Dishes Made With Mostly Specialty Fish
Fishing in many games is a pure time investment with a knowledge gate and occasionally a preparation gate (which is in itself almost always just more time investment). Supply therefore is 'infinite' relative to time and only needs calculation if a Clamp is involved. Risk depends on the game, or the fishing spot within the game. Demand is dependent on game type. UBI games use NPC demand, 10K games usually lead to the section title, fish dishes. These can still have some NPC demand depending on the game, again. You can increase Demand by increasing Dependence on this item, and in 10K games this automatically increases Reward, but Supply will rise to match it almost automatically. Final outcome is a profession of 'Seafood Cook' that is 'always in business' but can have 'slow days', and ofc all those Fishers who provide their fish, who are in a Competitive Loop with each other in the profession which is mildly decoupled from the Actual Demand. Not very easy for AI to handle.
Mushroom Dishes Made With Rare Mushrooms
Gathering Mushrooms is a time investment with competition at the source point in most games. Rare mushrooms are also moreso Luck than preparation in many. Supply may or may not be 'infinite'. Games where Mushroom mobs exist can keep it 'infinite but with a relatively known flow', while having some control over Risk (when the mushroom is gathered from a point in the world, control over Risk is not as clear). Demand matches the above example. Outcomes are 'Mushroom Cook' that is probably 'always in business' and Gatherers, depending on the game, sometimes 'Adventurers' or 'Dungeoneers' who specialize in Mushroom Gathering, but the Competitive aspect is clamped more directly by the amount of spawns available, meaning that Supply has at least a Rate Clamp (as opposed to fishing in most games other than BDO, where the Total Supply is unclamped (number of Fishers x Time investment gives a theoretically infinite amount of fish, infinite players would result in infinite fish, Infinite players does not result in infinite Rare Mushrooms). AI can do better here, if it is given some rough ballpark of 'total possible spawns (and droprates, if mobs are involved)'
Lamps And Similar Decorations
Making Lamps and similar decorations (we are counting 'similar decorations' as 'things players need more than one of, for a nice house, but almost never need any of, to get started, even if they care quite a bit) is a bit trickier, because there is no serious lever the designer can pull to force Dependence, nor do they want to, most of the time (TL actually tried a little, but it spreads out across all decor). A player will sometimes need food to do cutting edge content, but they will not need a wall lamp. Supply varies by game but is almost always tied to some explicit production skill, which may cut off some players from pursuing this and result in two things. First, the social aspect, why go through hours of work to make a Lamp when you can find a Lampmaker? Second, general enjoyment and pride in work for some, usually based on the other side of that interaction (but not always, NPCs could ask for lamps). Supply, though lower, might match either of the above. Demand is based on cost, and Reward therefore tends to cap out. NPCs requesting Lamps are almost certainly clamped in some way to prevent nonsense and botting. Obvious results are 'Lampmakers' and possibly 'Glassblowers', though most games have outcomes where one player does both things. Specialty lamps are where Adventurers tend to get involved. You can 'need 8 lamps' without 'needing 8 aweome/amazing lamps', and therefore Supply for the latter can be allowed to be much lower. Bargaining Power is the name of the game here, and AI therefore is less effective, basically being able to just do what it does for the Mushrooms at best, and write up some NPC stuff for 'why they want a lamp today'.
Cloaks and Similar Gear
Making Cloaks can be simple at low levels and complex later, but unlike lamps, in most games people only need one cloak, but often when they need it, they need it as soon as they can afford it. This causes most players of modern games to want this decoupled from any crafting, but I'm treating TL's success in this matter as a 'proof' that this isn't absolutely required (though their Trait system might be, since it changes the 'only needs one cloak' aspect). Supply therefore aims to always match Demand, a give-and-take that isn't quite the same as the Lamps. People might 'wait' for Lamps to be available when the Lampmaker is ready without contacting them to ask for regular Lamps. The social contract appears for Specialty Lamps. This is more likely to be the 'common occurrence' for Cloaks, except in games where leveling your Artisanship is much more grind than 'knowing material sources and having social interactions'. Basically if leveling Artisanship is too tedious, this tends to break. No one needs the number of cloaks that people have to make, to level. NPC demand often has to get involved. There is a pacing that can work out here (Guild Items and NPC Contracts work decently for the midlevels in FF11 and TL) but it depends on the material requirements and leveling speed in the game. AI effectiveness is low, it can't predict 'how many players will need this exact cloak enough to care' except for endgame cloaks and even then. Problems abound. AI ends up about as useful as for the Lamps, and the professions end up being 'Weaver', "Tailor', and ofc 'Adventurer' for those games where specialty cloaks are from mob drops. Gatherers are not counted here for the same reason as the Fishing/Mushrooms. They're too decoupled from the Demand to know with certainty what to do, and if the Reward is too high even with low Demand, someone will make a business out of negating the Risk. Low Demand makes this unlike the Mushrooms. Supply would always outstrip Demand so much in nearly any fun game, that NPC offers of reasonable Rewards will just make the whole thing crazy. One thing that helps is adding a different Supply clamp (FF11 Crystals, TL's 'Decoration/Furniture Scrolls' and 'Additives'), but then you're playing with fire at least a little bit.
Beds and Similar Decorations
The combination of the above two is the hardest. Players want beds but don't need them. Players only need one. No games I know of make you wear out and repair your bed's main function (BDO kinda). On the upside, this is great because no one can complain about RMT or P2W in the usual way. Even when not entirely cosmetic, the Reward for being a Bedmaker is tied to the health of the game/server, or to NPC requests for Beds, which generally would be rare for higher quality ones. Beds are often the most demanding in terms of total materials, in games, even if not requiring a huge amount of 'Artisan skill at Bedmaking', and almost never 'a direct path to leveling your Artisanship'. Specialty beds often require Adventuring to be involved, to avoid it just being way too luck dependent, but it can also be 'just a grind'. Beware of this though, because some players are truly obsessive. Often without reason (we have seen a screenshot of a player whose house contained multiple Gigantrite/Minutrite beds, there is no benefit to this, we almost hope this was just a proof of ability to dupe things or something). Bargaining Power is weaker here, unless something forces Bedmakers to be even rarer (in FF11 for instance, it wasn't enough to just be a Woodworker, to make Beds you also needed to level Clothcrafting quite high for the better wooden ones, without the combination skill, you couldn't succeed). AI can do almost nothing for design here, this type of item is almost entirely a function of the social health of your game, because Demand is low, and without involving mob drops from Adventurers, the designer has almost no control over Risk or Supply, twisting this into an item where economic decisions are made entirely based on Obstacles vs Reward, and if nothing else from this section is memorable, at least remember this… in a reasonable MMORPG, unless we are talking about 'did this boss drop the thing', Obstacles Aren't Real.
I miss my text colors and spoiler tags a lot this week.