#and interpret that creation. Even if you do admire AI as a concept you probably admire how the idea reflects those human connections

[vampthropologist]

Companies use AI the same way companies run off of a quantity over quality system. If a product is passable enough to make a profit, they sell as much of it as possible. If an AI is just decent enough to string sentences or make an okay image, they sell subscriptions to it as much as possible and generate as much content with it as possible.

Social media is built off of posts, so they get AI making posts. They make those posts so fast in such large amounts, and even if they are shit it brings revenue. Instead of people taking time to create and post that online, they just turn out whatever is halfway decent with an AI and rake in the money. It draws attention away from creators, because in the corporation's mind, attention is time is money and if you're looking at something it benefits them. If they shove a bunch of AI in your face, it might flush the creator you like further down your timeline or For You Page but makes the company a profit.

Quantity > Quality, except moving from consuming objects to consuming your time and attention (even more than they already have, given the "attention economy" and all).

[symbianosgames]

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community. The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.

If you've played 4X games (eXpand, eXplore, eXploit, eXterminate), you're probably acquainted with this scenario: late in the game, you've built enough population, researched enough technologies, and constructed enough buildings and units that you have a massive advantage over all remaining players... but the game still makes you trudge through many more turns to get to an ending.

Sound familiar?

The Silicoid player will not become the next Master of Orion.

This is the experience most people will have when they play a strategy game. So to the extent that 4X games tend to be varieties of strategy games, they elicit a similar "I won, why am I still playing?" reaction.

I'd like to suggest that the essential problem here is not end-game micromanagement. That's the visible symptom. The underlying problem, I believe, is actually inherent in the nature of strategy itself: a good strategy wins (or loses) long before you find out whether you've actually won (or lost). The rest is just waiting to see this finally happen.

Waiting, generally, is not fun.

Fortunately, since these are games and not reality, it's possible to design strategy games (including 4Xs) so that players who like this sort of thing can skip the boring bits. This means it's important to understand what "strategy" actually is, otherwise 4X games will continue to be less fun than they could be.

There's a well-known model in military science that, broadened a little, applies well to any large-scale goal-oriented organization. My version of this model, refined for strategy gaming, looks like this:

Grand Strategy (Vision: holistic thinking, applies to whole organization)

Strategy (Planning: deep thinking, applies to large areas over long periods of time)

Operations (Logistics: detailed thinking, applies to regions over medium time)

Tactics (Action: fast thinking, applies locally and immediately)

Each of these levels is composed of multiple instances of the levels below it. An operation, for example, will be accomplished by multiple coordinated tactical engagements.

The nature of strategy -- since we're talking about 4X games as strategy games -- is that it it rewards thinking carefully and creatively about how forces change over space and time, and it applies over large areas and long spans of time. You can set a strategic plan into action, to be carried out at the operational and tactical levels, but because all the individual results at those levels must accumulate over time into strategic information, it takes a while before you find out if your strategy was good or bad.

Seen in this light, the problem with strategic games, including most 4X games, becomes clear. Once you're at about the 2/3 or 3/4 mark in 4X games as typically designed, your final strategy has been set in motion. At that point, for all intents and purposes, the game is over. Your strategy, as it gets applied, has already either won or lost the game for you. Now you're only whacking the "Next Turn" button repeatedly to see how you won or lost. And that can't help but feel like pointless busy-work. It's an unsatisfying ending, especially after spending many hours exploring, expanding, and exploiting.

This problem is compounded by 4X games that define the core gameplay loop as frequent management of units, perhaps in the belief that "players will get bored if we don't give them something that needs to be done every ten seconds." At this point you have what says it's a strategic game, which should elicit and reward deep, big-picture thinking, but whose actual mechanics demand and reward quick, focused thinking. There's nothing wrong with quick, focused thinking as gameplay -- it's just a mismatch if what you're trying to craft is a strategy game.

As a result, players who feel something is off are correct. They are recognizing an unresolved game design conflict, wherein the tactical-level, action-oriented mechanics implemented for most 4X games actually block the player from experiencing the strategic-level, perceiving/planning fun that a true strategy game must deliver.

Even the most highly-regarded 4X-style games, such as Civilization and Master of Orion 2, fall victim to this. The last stage of these games, like other 4X games, is just button-mashing mop-up. You know who's going to win. And yet your last active experience in the game is just finger calisthenics. Boring!

This appears to be a hard problem in game design. How do you deliver actual strategic fun without resorting to high-frequency but lower-level mechanics that become tedious by the endgame? No developer of whom I'm aware has figured out a good solution to this problem yet. (Proposed counter-examples to this opinion are welcome.)

So we keep getting 4X games that either devolve into pointless end-game grindfests, or else have some feature bolted on that causes them to stop being true strategy games. (I'm looking at you, real-time clocks.) And we wonder why 4X games aren't more popular.

I don't pretend to have The Answer to this problem myself. I haven't had the chance to actually implement any of these notions to see if they work or not.

Bearing that in mind, I believe the following design suggestions would lead to more satisfying 4X games:

1. Get rid of individual units. The moment you include individual player-movable units as a feature, even if you "chunk" them to a strategic level (e.g., one object represents an army or a fleet), the temptation is nigh irresistible to make the core gameplay loop all about the player moving those individual units around on the map every single turn... and that is how you get endgame micromanagement grind.

In a truly strategic game, the most you would do is provide goal guidance to administrators and general officers: "Admiral, I expect you to secure these three key resource centers within the month," or "Doctor, this agency must produce one breakthrough technology in both of these two sectors by month's end." Note that the key concept here is delegation. This implies AI for translating goals to projects and projects to actions.

2. Make gameplay choices continuous, not discrete. In other words, the actions the player performs most frequently in a strategic game should not be a hundred repetitions of "do this one specific thing," but rather "continuously seek to accomplish this general goal until it's completed, impossible, or I tell you to stop." Another way of saying this: let the player manage logistics at a high level. Even if the player doesn't move units, it's a necessity for strategic (and thus 4X) games that logistical choices must matter. Minimization of logistical cost is half of strategy. (Maximization of resource value is the other half.)

Players need to be able to tell the game "move this strategic resource from this source to these destinations," and for the game to keep doing this on the player's behalf until it no longer needs to, or it can't, or until the player tells it to stop. The high-level commitment of resources as a one-way and continuous choice that can't be altered immediately makes strategic choices matter. So when the player directs the distribution of a particular strategic resource from its creation point to its usage points, it has to take time for this movement to happen, and it needs to apply continuously over time.

These features let strategic decision-making feel substantively different from the much more frequent and immediately responsive tactical gameplay. (I'm aware of the counter-argument that Master of Orion 3 was unpopular because it tried to abstract away micromanagement. My rebuttal: had it tried harder, and been called something else, it might have been more successful.)

3. Include personnel management. This isn't as crucial as the above two suggestions, but it's valuable enough that it's worth considering. Because strategic play choices are carried out indirectly and have consequences over time, there's room for the personalities of leaders to play a useful and fun role in accomplishing the player's strategic goals. (This may make more sense when you think of NPC leaders as a strategic resource.)

Leader personalities can add intriguing variation to how a player-assigned strategic goal is interpreted and accomplished. Thus, gameplay mechanics for grooming and selecting NPC leaders, and for letting the player manage conflicts among leaders on how best to accomplish goals, can to some extent compensate for removing the ability to control individual units.

4. Opposing strategies should be perceptible and plausible. In addition to the internal challenge of making effective choices regarding finite resources, a strategist also faces the external challenge of competing strategies intended to accomplish other results. A 4X game needs other players, human or AI, whose goals oppose the player's in some way. This usually isn't too hard; what's not always done well is reflecting these opposing strategies to the player.

This doesn't mean explicitly telling the player exactly what an opponent's strategy is. Part of the fun of strategy in a game is trying to identify an opponent's large-scale intentions based on observing the individual actions taken by the opponent's agents in the world and perceiving a pattern to those actions.

In a game, these patterns need to exist as an opposing player's intent; they need to be detectable so that the player has data from which to construct a model of an opponent's intent; and these intentions need to seem rationally plausible as a plan for advancing the opponent's interests and not merely random, irrelevant activity. A good strategy/4X game will be designed so that opponents have a strategy, their gameplay choices seek (like the player's) to execute this strategy effectively, and, given data and time and perceptiveness, it is possible for the typical player to figure out an opponent's strategy so that it can be countered.

Until it changes, creating a new puzzle to solve.

5. Know when the game is over. The engine of a good strategy game continuously assesses the state of the game world versus the selected victory conditions and detects when further player choices will have a negligible effect on the final outcome.

There's more than one way to report this information to the player. The game may may show the player a running calculation of the odds of strategic success. Or it may hide this information as a strategic version of the "fog of war," encouraging players to make this calculation for themselves.

Either way, once this internally calculated "chance of victory" value reaches a certain point, the game should announce that the contest is all but won or lost, then give the player the option to stop playing or to keep going to see the final state of the game world. It may even offer to simulate some additional turns until a stable state is reached. The important thing is that the game itself recognizes when the victory conditions are satisfied or unsatisfiable, and it doesn't mindlessly ask the player to keep performing mechanical actions that no longer have any meaningful strategic consequences.

Would these suggestions for emphasizing large-scale exploration, pattern-perceiving, and planning deliver more of the particular kind of fun that players of 4X games are looking for?

There's only one way to find out for sure.