The truth about AI content generation

Henrik Roth
6 min readMay 28, 2021

From chiselling into stone, scribbling on papyrus, writing with ink on paper to technologies like the letterpress and internet. Technology has always been supporting copywriters and marketeers. Today, we even got to the point where computers write our texts. Copywriting software has been around for a long time, but what technology did OpenAI bring to the market with GPT-3 in 2020 and how will it change the job of marketeers and copywriters?

Let me introduce: AI

The past few years there has been a big hype about AI. But what exactly is AI? AI is short for “artificial intelligence” — the development of machines to make them smart so that they carry out human tasks such as problem-solving and decision-making. It includes “machine learning” and “deep learning” — two names that often are used interchangeably. So what is the difference between them? Let’s put once and for all an end to this question. Both are subgroups of artificial intelligence, while deep learning is a subgroup of machine learning. Machine learning is just what it sounds like — a machine that is learning. Just like kids during their school time, a computer system learns and adapts without a human giving it any kind of instruction. It uses algorithms to analyze data, to create a pattern and to then use it for other tasks. Deep learning is a large neural network — the “brain” of a machine. It tries to imitate the human brain by processing data and based on that creates a pattern to use for other tasks.

GPT-3 — What’s the deal about it?

2020 — the year of birth, the year that changed the Tech-world because GPT-3 was born. GPT-3 (Generative Pre-trained Transformer 3), launched by OpenAI, is an autoregressive language model that uses machine learning and deep learning. It is based on a gigantic neural network with 175 Million parameters, which is a lot, considering a human brain has 125 trillion synapses. It is the biggest Natural Language Processing Transformer (NLP) which cannot only write and complete texts, but it can do it in a human thinking and reasoning way. GPT-3 is even as far as writing texts that are written so well, that most readers cannot distinguish them from human copywriters.

GPT-3 uses AI algorithms and already existing content in order to create new content. The question is: how does it do it? The AI collects data, looks at certain patterns and tries to reach conclusions from it. It then gets feedback in order to improve its algorithm. When the algorithm is good enough, the GPT-3 based AI is able to use its knowledge for new data. The important thing here is that it is constantly being “fed” with large data in order to improve its algorithm.

What can GPT-3 based AI do?

From your personal writer, to an investigative journalist. GPT-3 is very versatile. What else can it do? Here are a few examples from an endless list:

  • Generating codes and reading written codes
  • Designing website templates
  • Showing how to generate a JSX layout using simple language
  • Generating regex
  • Cloning websites
  • Object use-case generation: you type an object (e.g. rice) and it tells you what you can do with it (fry, cook, bake, grind)
  • Creating diagrams and tables using simple language
  • Playing games or evaluating objects (e.g. Art pieces)
  • Philosophizing
  • Writing machine learning models
  • Generating formulas
  • Creating animations
  • Creating 3D scenes/objects
  • Creating resumes
  • Generating live sports tickers
  • translating languages

Benefits of AI content generation in Content Marketing

  1. Creating creative content: Human copywriters do not have to fear writer’s block or creativity pressure any more. AI software can help by generating various creative content to choose from.
  2. Speeding up writing: “time is money” — we have all heard about it and the importance of it makes many copywriters feel immense pressure. However, with AI generated software, they do not have to fear no more! AI generates many versions in split seconds — only with a touch of a button. This saves many writers essential time.
  3. Aiming at target audience: most marketeers know the importance of addressing the target audience. Many years ago they had to make sense of it themselves by surfing through the internet, doing surveys, and going through other kinds of media. Now they have AI, a machine that is more precise than the human mind and is able to use information from each customer to create content specifically for that individual and address it towards her or him.
  4. Validating performance: AI does not only create, it also validates. After generating great content for copywriters some AI powered software are able to tell copywriters the predicted result of each copy. For example, some tools generate newsletter subject lines for marketers and afterwards tell them by how much their open rate increases, if they would choose that particular subject line. This is what neuroflash’s tool does. A few keywords or a descriptive text and with the click of a button it generates various subject lines with the predictions of the AI.
  5. Being very versatile
    Whether AI is being used for e-mail marketing or whether it is being used for writing marketing copy — AI content generation supports whatever. AI is able to complete many tasks, like introduced earlier. Following examples are based on content marketing:
  • Creating Newsletter subject line
  • Generating app push notification
  • Writing product descriptions
  • Producing headlines
  • Creating marketing copies
  • Keyword finder
  • Research topic suggestion
  • Finding images

“A robot wrote this entire article. Are you scared yet, human?”

This is the headline of a famous article published by the Guardian and completely written by an AI. The question is legitimate because many writers now fear for their job. Do copywriters now have to die a slow death? Are machines going to take over their jobs? Is this a battle between humans and AI? A few years ago when AI already existed few people were really concerned — many even laughed about it. They did not think that a machine is trustworthy enough to take over human jobs. Back then they were right because AIs really was not trustworthy enough and just not “smart” enough. That has changed now. New technology, such as GPT-3, turned AI into being more woke, human-like, and ethical.

However, human writers do not have to fear for their life yet. AI is smart and can in fact create very good content, but what it is missing is human life experience, knowledge, and most importantly emotions. It can be very rational and objective, but it cannot create emotions. AI content generation tools are here to help humans finish their tasks faster and support them with creative ideas — not to take over humanity. Like Bill Gates said “AI can be our friend”.

The future is now

In 2015 10% of businesses have used AI-technology. In 2019 the number rose to 37%. That means that in four years the number of businesses that adopted AI grew by 270%.(Source: https://www.gartner.com/en/newsroom/press-releases/2019-01-21-gartner-survey-shows-37-percent-of-organizations-have) Many people were (and many still are) suspicions towards AI. They do not think that a machine can be accurate and that it can be trusted. However, the truth is, that many businesses cannot live without it, and that it is the reason for many businesses gaining in profit. It has been predicted that by 2022 30% of digital content will be generated with the help of AI-based technology and that by 2027 the AI market is expected to reach $267 Billion. (Source: https://www.fortunebusinessinsights.com/industry-reports/artificial-intelligence-market-100114)

The development of AI never sleeps. Experts have predicted that instead of searching all over Google, there will be virtual assistants that take your personal likes, your emotions, your habits, and your reactions into account in order to recommend you specific content. Also, content creation is right now limited to mostly short to medium texts. In the future AI generations tools are predicted to compose long chapters or even whole books. (Source: https://www.zdnet.com/article/ai-powered-virtual-assistants-and-future-of-work/)

Artificial Intelligence is not the future — it is already in our lives without us fully realizing it. While being very beneficial for many businesses and people, still many people are afraid using it. Don’t be!

--

--

Henrik Roth
0 Followers

Passionate entrepreneur with over 14 years of experience in digital marketing, entrepreneurship & growth strategies for start-ups. CMO & Co-Founder @neuroflash.