Posted by Computer-Assisted Content Advancement
Consumption: Computer-Assisted Content Advancement
Video, image, and game-sharing platforms have cemented themselves as a central part of the general consumer’s internet experience. With the sheer volume of content uploaded to platforms like YouTube, Twitch, Reddit, and Newgrounds , it is difficult to truly say there is nothing on the internet that interests you; even the most niche interests are bound to be represented in such a large mass of content. Consumption can be defined as the adoption of a product by the consumer and making it a part of their life. We see this in the media we consume and the platforms we adopt online, embracing the identity that comes with being a user of a certain app or a community member of a particular group of consumers.
With the advent of the internet there are so many avenues to getting content in front of a large audience. However, there is a barrier of entry to quality content creation. Quality content has been refined by the producer with skills such as video and image editing, scriptwriting, and artistic talent to create a product consumers will want to engage with. Learning these skills and polishing that craft can take a large amount of time for the producers. As a result, we as consumers are left with piles of low-quality content to sift through in order to find high-quality media that appeal to us.
In the past decades, we have seen major advancements in the field of artificial intelligence (AI). The versatility of AI networks is undeniable. Thus far they have been successful in performing and replicating human tasks, being trained to do everything from playing chess to understanding human language. Based on these past experiments and the current progression of AI, it is fair to believe that we are on track to developing AI that could assist in, perhaps even create, online content meant for human consumption. I will refer to this AI-assisted content production as “Computer-Assisted Content Advancement (CACA).”
I would like to focus on a few applications of AI that I have discovered that could assist in content creation. First, we will look at AI image generators. Dall-E and Craiyon, created by the research lab OpenAI, are two of the most popular models. These programs take text inputs from a user to generate images recreating their interpretation of the given prompt. A consumer can simply type in what they want to see and get it back in minutes with no need to seek a creator. AI can also be used to create complete scripts and narratives . Playground, also created by OpenAI, takes a user’s input and then writes a script based off it. A user can tell the program to create a script, story, tagline, etc. The text generated by this program is often short and rudimentary, but it serves as a starting point to create a full script off for a creator. A more advanced application of AI-text generation is AI Dungeon. This videogame uses an AI to create text-based adventures from user inputs. Even though the AI is being used in this context as a game tool, it could be applied to other writing tasks, such as producing articles or writing scripts for content creators. There is also AI that generate music in a multitude of genres. Instead of searching for a good song, a consumer could simply request the AI to make one for them. Whether a creator needs music for a game, a video, or any other project, there are AI engines that can produce them. This negates the need for a content creator to learn musical composition or pay a musician to do it for them. One standout example of this is AIVA, an AI classical composer that is so advanced it has been recognized as a composer by the Sociétédes auteurs, compositeurs éditeurs de music (SACEM), a high-profile classical music publisher.
Images, scripts, and music can be generated by AI programs readily available to the public for free, so why isn’t everyone using them to assist in creation for their own consumption? There are two major limitations to this application right now: intelligence and cohesion. Because these are new developments, AI applications still have much progress to make. AI require massive amounts of processing power to function at higher levels on par with human cognition, something which we have not developed yet. Applications like Dall-E and Craiyon, while impressive, still cannot reliably generate images on par with a skilled artist’s creations. Additionally, these applications are all separate programs. For a person to utilize all these AI programs in CACA, they would have to run all these applications separately. Furthermore, while a creator may provide all these applications with the same input each of these networks are different for many reasons, such as algorithm or processing differences. This creates variation in the output of each individual program, leading to incoherent material the creator must piece together into consumable media. In its current state, using AI to assist in content creation is a fun novelty, but largely impractical when considering wide appeal to general consumers. Perhaps in the future as these applications continue to be developed and consolidated, they will manage to surpass human capabilities, replacing people as the primary content generators for general consumers.
Computer-Assisted Content Advancement could vastly streamline the process of online media consumption. No longer would the consumer have to search far and wide for good content that appeals to them. By lowering the barrier of entry for good content creation through assisting the producers at little to no cost, more quality content of a larger variety will be available to consume. However, the ramifications of employing CACA are concerning. At the core of consumption on the internet is the variety of people creating content and communities to join. There are so many different groups represented on the internet and often people become introduced to them through consuming media they do not normally engage with. With more content available that actively caters to each individual consumer’s interests, people may be discouraged from venturing out of their niches and discovering new interests. This enriching experience of discovery innate to the internet is lost when the consumer has no need to discover. Finally, if AI does surpass human content creators this could spell the end for many creative jobs. Why would a creator pay a scriptwriter, an artist, a musician, or an editor to help them in content creation when they could get similar work from a single AI unit? Why would a consumer pay for this content when an AI network could make it for cheaper? The automation of these jobs could displace many creative-types, leaving them structurally unemployed. The prospect of CACA is an exciting one for the consumer looking for more appealing content, but perhaps frightening for the creator that has spent so much time polishing their craft.