Games
Game building is an interesting niche of coding. Almost anyone can make a video game, assuming they have access to a computer. They might even make a good one if given a few resources! Video games can be huge, sprawling projects that take the player through entire universes, or simple, small, platforming games that end when you rescue a princess.
For a brief period of time, a person could tell which one of these games was which by file size; before, there was a hardware limitation that meant any game for a given year was under X MB or X GB for a given system, because the cartridges or the disks could only hold that much, and game designers had enough time to properly use all of it no matter what kind of game it was. Sonic, for instance, famously dedicated an absurd ratio of its cartridge space to the Sega! intro cry, but it had to do that because there was no other option. It cost them space they could have crammed more assets into to properly brand their game!
After this period, games were all sorts of shapes and sizes, and games purchased online took off. Games were unbounded by the requirement to fit on a given disk, and instead bounded by the hardware of the device they were playing on. In online games, this meant they’d look a bit worse than most console games, but the size of the file was then a direct ratio to the actual number of levels, lines of dialogue, character models, and other assorted assets within the game. The days of DOOM, a beloved game that to this day is still playable on basically any hardware (including hardware it’s not supposed to be played on like calculators and especially robust microscopes) were over – games no longer had a double-digit megabyte ceiling.
Then, after that, things started to go awry. Graphics started getting better, and audio fidelity improved. The question became not ‘how well could an average TV play this audio file?” but “how can we make sure the people with 300$ headphones are pleased with the audio quality?”. Code, too, started to sprawl out. Where DOOM is airtight, games like Fallout 4 or Borderlands 3 assume a powerful computer and a good graphics card, so the code can be inefficient at points if it means they’ll get the game out in time for Christmas. The devs genuinely do not have time to make everything as air-tight as DOOM anyway, and in fact such a thing may not even be possible when games moved from 2-D to 3-D representations of the world!
Good games and bad games alike are both equally subject to a buggy launch day. It’s honestly a failure of modern day game design that first-day launches are just assumed to have bugs. Worse is a market that lets them by buying things as soon as they’re available, not as soon as they’re good, when that information has never been more available.
Work Programs
Many work programs are suffering similarly under the sudden gold rush of generative AI. Whether the product is ultimately better for having it or not, AI assistants have been introduced just about everywhere as quickly as possible. That is to say a largely experimental feature has been introduced at incredible speed to programs that were otherwise totally functional and stable beforehand. Even the best of the best, the flagship examples of apps with AI, struggle to find use for them that forums don’t provide.
Does Excel need CoPilot when it already had Pivot Tables? In fact, is the CoPilot going to give me any shortcuts to making Pivot Tables, the single thing I find the most difficult about Excel when trying to categorize large amounts of data? I think so – but it seems like while CoPilot can fill individual boxes and manage formulas for me, the PivotTable with the items I wanted organized in the way I wanted, plugged into CoPilot with simple English, is a little beyond its grasp. In fact, while trying to test it at the time of this writing, it seems to have crashed! I have a beige panel I can’t exit out of or clear. The ‘Analyze Data’ tab, which I would guess is running off of a similar AI but more closely tailored to Excel’s specific needs, offers to make a Pivot Table out of highlighted data by default, albeit one that I would have to rearrange.
Adobe, too, has introduced an AI companion, one which is supposed to make Smart Replacement (where you lasso-tool a particular object and tell Adobe ‘replace this with another object’ or ‘erase this’) a reality. Smart replacement is really cool and sharp when it works, but it takes a lot of data to improve; Adobe has a vague handwave suggesting users give them permission to use their content made in Adobe as training (https://www.howtogeek.com/858952/adobe-is-using-your-data-to-train-ai-how-to-turn-it-off/). That makes sense, right? But, ultimately, customers don’t really like paying to be the guinea pig for new features if they didn’t opt into that. It’s why you’ll see certain websites that are trying to improve the average user’s experience offer an opt-in to their development pipeline, so they get their crucial data, and users who didn’t opt in don’t have to contend with the app suddenly failing, crashing, offering new things that don’t work quite right, et cetera. In gaming terms, this is the ‘Early Access’ version, which is understood to be incomplete. Beta testers are another option, but hiring testers is expensive and takes time. Once the app is stable, it is then generally launched to the wider audience. Adobe did some testing, certainly – and then, for the sake of speed, launched the Early Access and figured they’d patch it on the way. Whether it was ready or not, it was out on the market as soon as other apps began offering it as an exclusive service.
Art Programs
Digital art, the process of using digital tools to create art, should not be confused with art created by Generative AI. On most art platforms, calling one the other is a surefire way to pick fights. Digital art can be made in many, many ways, and includes everything from digital collage to digital painting to digital photobashing and more. If there’s a real-life form of art, chances are there’s a digital version. Even things like sculpting can be done digitally in programs like Blender.
Some artists sheepishly admit they do all of their digital work on their phone. Not a tablet. Not the computer. Their phone, with its touchscreen, and sometimes their fingers instead of any sort of stylus. This is the sort of revelation that makes people sitting at the middle of the bell curve of art stare blankly into space for a minute – what do you mean you make high-resolution art on your phone?? They get this reaction for a couple of reasons, firstly because it is more difficult, but secondly, because art programs are getting kind of annoyingly huge!
Far from the premier programs like PhotoShop and Adobe, Clip Studio Art and IbisPaint offer a pared-down experience that means the app can run on basically any machine including a slightly older smartphone. Here’s a problem: the reality of top-of-the-line tools is that they take top-of-the-line computers to run them. This is, once again, a sticking point for these products – what are they doing that they take an Nvidia 3600 to run without lagging or crashing, when a hypothetically equivalent action in another program doesn’t? How can it be possible that some artists can make gorgeous, expressive, brilliant art on their phones, and at the same time a kid trying to run Photoshop on the family computer can’t get it to draw a straight line for her without lagging?
The art programs designed for phones are designed not to use many resources, for better or worse. The art programs designed for computers, are assuming that, like video games, they’ll have a user with the ideal machine who will benefit from a crazy high resolution, and anyone else is an edge case who’s probably not going to ask for a refund anyway.
The Takeaway
Programmers are being run ragged trying to meet demands made by people who think they know what the customer wants. And to be sure, they know what some customers want, but inevitably they’ll be shorting the remainder more than they’d be benefitting the people with the best computers. Tighter programming that eats fewer resources while running does not only benefit the people trying to work off of smartphones or off of computers that are already five years old – trying to write a game, or art program, or any program well enough that it works at launch with minimal bug fixes used to be the norm because it saves work and stress on the developer side too. Meanwhile, programs that are written assuming the computer will rise to meet its needs instead of the other way around often create products that – in a vacuum – should have worked but instead would take resources that don’t even exist on the market yet to properly run.
If you need help procuring software for your business that won’t fry your hardware, or better hardware to meet compliance standard software, get in touch – we can help! https://www.elixistechnology.com
