Behold! A new decade is upon us, which means it’s prime time for prognostication. What will learning games look like in 2030? What does the future hold for our stalwart, frontier industry? Welcome to Part III of Learning Games in 2030. If you missed Part I (survivability and mixed reality) or Part II (robot-era skills and esports), fear not, you can find those articles here and here, respectively.
Let’s dive right in with prediction number five: in 2030, learning games will be streamed. Remember OnLive? Launched in 2010, OnLive was an early cloud gaming service that was fatally ahead of its time (OnLive closed its doors in 2015). While still in its infancy, game streaming has been resurrected by platforms like Google Stadia and GeForce, which are investing heavily in the technology. So what is cloud gaming, exactly? Simply put, it’s a service that allows players to play video games on remote hardware. The game software is “stored, executed, and rendered on the remote…server” and the video results are streamed to the player’s computer.
A key benefit of this technology is that it “largely makes the capability of the user’s computer unimportant,” which is HUGE for games in schools. Currently, games need to be browser-based in order to be widely adopted by schools; not because web browsers are an ideal platform for games (they’re not), but because they’re the lowest common denominator. Browser-based games work on most devices commonly found in schools, including Chromebooks, which accounted for more than half the mobile devices shipped to schools in 2017.
Don’t get me wrong; web browsers are a perfectly viable way to deliver quality learning game experiences. That said, many outstanding learning games are not – and never will be – browser-compatible, and that’s unfortunate. If freed from the constraints of the browser, learning games could have rich 3D graphics, high-fidelity physics, and, in general, much higher production values (on par with their entertainment-oriented counterparts, even). For instance, the robotics game RoboCo we’re currently developing is too graphically rich (read: demanding) to run on low-end hardware like Chromebooks. If cloud gaming takes off and schools adopt it – a very real possibility by 2030 – students will be able to play high-end games like RoboCo on any device with a good internet connection.
Speaking of high-end games, prediction number six is that learning games will be nearly indistinguishable from entertainment games. As we’ve discussed, schools will either have high-end gaming hardware via esports (see Part II) or be able to run high-end games on low-end hardware via streaming. Either way, they will be equipped to run high-end games in 2030; graphically rich, meaty games that look and feel like the games people play for fun.
It’s currently very economically challenging to develop a high-end game for the school market because the production costs are exorbitant and the typical school’s budget for purchasing games is low. As a result, many learning game developers – including Filament Games, Strange Loop Games, We Want to Know, and Ululab – are attempting to make their proverbial nut by targeting consumer entertainment and/or parent customers before investing in institutional markets. I expect this trend to continue until schools at large have an appetite for games on par with other forms of media.
In some respects, this trajectory represents a natural evolution of the learning game genre. In the early years (the 80s – 90s), learning or “edutainment” games were often awkward combinations of unapologetically rote tasks (e.g. multiplication) and academically-irrelevant game mechanics. Though these types of games are still commonplace, the 00s and 10s saw the rise of robust contemporary learning games like those found at Sandra Day O’Connor’s iCivics and Robert Ballard’s The JASON Project (now Jason Learning). Though still not sufficiently robust to compete on mainstream gaming platforms like Steam, these games featured deep(er) systems, (more) complex interactions, interesting choices, and the opportunity to experiment with different identities. In the 20s, I believe learning games will continue to evolve toward consumer entertainment games in terms of both gameplay quality and production values, and by 2030, will be held to the same standards.
Finally, because I can’t write an article about the future without mentioning “AI”, prediction number seven is that learning games will leverage machine learning to deliver custom scaffolding. Non-player companions that effectively serve as tutors – for instance, Half-Life 2’s Alyx character – are commonplace in games, they’re just not particularly intelligent. As machine learning algorithms improve, these companion characters will gain the ability to observe players’ in-game actions and offer increasingly unscripted, useful, and individualized assistance. Will they replace real teachers as Marc Prensky predicted in his book ”Don’t Bother Me Mom – I’m Learning” to any substantive extent? Not by 2030. IMO, longitudinal human observation and natural dialogue will trump automated assessment instruments for decades to come.