The phrase “artificial intelligence” gets thrown around in the tech and marketing world a lot. It’s used as a buzzword similar to over-used language such as “data-driven,” “optimized” or “growth hacking.” Phrases like those raise our collective hackles unless they’re weighted with the truth.
Announcements of game-changers in the space are often dressed up with smoke demonstrations showcasing cherry-picked use cases. Such sales presentations often paint a vaguely inspirational picture of how a new product, or piece of technology, is going to disrupt the industry. Remember Theranos? Anyone?
We don’t take it lightly when we say the recent advancements made in AI are a big deal. Let’s break down why:
So, what is AI?
Artificial intelligence is defined as: “A branch of computer science dealing with the simulation of intelligent behavior in computers…” AI takes a set of outside inputs and provides predictive modeling to forecast and determine the outputs — constantly refining as it learns.
In layman’s terms, the computer is guessing, checking, and revising, as we were taught to do in first grade. The power comes with how quickly and accurately a computer can do these things.
American culture has been obsessed with the idea of AI since it began to crop up in novels in the 1950s, and as such, when people hear AI, they think of what is known as generalized AI. Generalized AI is an incredibly flexible AI that would be able to imitate a human brain (think Hal9000 from A Space Odyssey).
Technology hasn’t gotten quite there, but we’re getting closer. Enter stage left, GPT-3.
What is GPT-3?
GPT-3 was created by OpenAI, an artificial intelligence research lab, and will be accessible to anyone. GPT-3 is short for “Generative Pre-training Transformer” and is the third iteration of the machine learning model of its kind. There are many different types of artificial intelligence use cases, and the blanket term can be misleading when discussing technological applications.
It’s not all big-picture thinking like self-driving cars and computers that could someday pass the Turing test. Every internet user interacts with some type of AI daily through everything from social media, to how we’re advertised to, and even our email spam filter.
Among the ubiquitous examples of machine learning-driven, AI is Google search. With hundreds of millions of new domains every year, Google is not manually indexing and deciding every possible search result beforehand. Their algorithm helps determine content, relevance, and whether or not to display a particular page. An honest to goodness example of machine learning.
GPT-3 has many different capabilities and it is by no means perfect or an immediately viable solution. Kevin Lacker, a computer scientist, and former Facebook and Google engineer, actually did run it through a Turing test and found it’s quite good at common sense-type prompts but still subhuman in other areas.
How could GPT-3 change the world of marketing?
Understanding the potential use cases for where we’re seeing early adopters start to leverage this tech leads us to hypothesize how this can shake up the world of marketing in particular.
- Web and App Design & Development
Here we see two curious use cases that could have big implications. Could GPT-3 potentially streamline both design work and web development to remove some of the brass tacks back-end work and dedicate more of their time on tech advancement?
We could see a future with the opportunity to spend more time focused on strategizing and ideation and the ability to activate big ideas into functional applications in seconds.
- Content Marketing
When it comes to content marketing for purposes of speaking directly to your audience, as well as gaining ranks in search engines, one of the biggest bottlenecks is the actual production of the content. Only so many sentences can be written in a workweek by any individual person or team.
Check out this site that had GPT-3 write pages on pages of creative fiction. They’ve even got accurate parodies of popular copypasta and Harry Potter chapters that never existed. It’s not half bad.
True synthetic content that people find difficult to distinguish from the human-written text could point to future automation processes in the realm of content marketing.
We can think of specific use cases where ongoing and never-ceasing content generation could be particularly useful for pedestrian tasks such as Google text ad generation or social media community management responses.
- Marketing Strategy Development
Seeing GPT-3 build a pretty coherent and strategically directional marketing strategy are milestones themselves. Can we leverage AI to amplify our thinking and find whitespace in future opportunities for our advertising clients?
We foresee the opportunity to tie in this technology when it comes to reading and interpreting data. Can we train the AI to provide valuable insights into campaign performance and make continuous improvement recommendations? There are many implications on how we can see this tying in with marketing as a whole.
Announcements of game-changers in the tech space are often dressed up with demonstrations that showcase cherry-picked use cases. But GPT-3 — with many different capabilities — could be an enormous shift in how AI could shake up digital marketing.
Rebel stays focused on these changes to bring the latest and greatest to our clients. Learn more about our Emerging Technology practice.