Artificial Intelligence in Marketing: GPT-3, AI Writing & PR Tools

The Ultimate Guide to Creating a PR & Marketing AI Strategy 

 

The global writing assistant software industry is expected to be worth $763.93 million in 2026.

A new industry exists around optimizing digital content for comprehensiveness on Google search results.

Every business needs content to survive in the digital economy. But creating that content can often be a challenge for business owners with limited resources.

Enter the world of copywriting tools using artificial intelligence (AI), Natural Language Generation (NLG), and the Third Generation Generative Pre-Trained Transformer (GPT-3).

AI and marketing automation tools can drive tangible business growth by creating content discovery opportunities on Google and in Web 3 communities.

AI copy tools can also help you generate additional search traffic with robust AI-generated content that would otherwise cost thousands in Google Ads.

In this article on the use of artificial intelligence and machine learning in digital marketing we discuss:

  • Can Artificial Intelligence Help Your Brand Marketing?
  • How is AI used in B2B Marketing?
  • How can you create an AI PR & Marketing Strategy for your business?
  • The best AI copywriting tools
  • How to generate marketing, advertising & PR content with AI

Decoding the knowledge graph: GPT-3 and Content Creation

AI, NLG, and GPT-3 copywriting tools.

AI copywriting tools can help content marketers, search engine marketers, and writers understand how to:

  • Reverse engineer how Google thinks about concepts
  • Connect the dots on Google with semantic concepts

AI copy tools and semantic SEO AI tools help marketers build a knowledge graph and boost organic search traffic.

Ranking on Google entails covering topics comprehensively. But how can you possibly know how to cover a topic comprehensively unless you have analyzed thousands of data points and metrics? AI copywriting tools are used to assist with this process.

According to The Marketing AI Institute 2021 State of Marketing AI Report, 41 percent of marketers are achieving revenue acceleration with AI today.

So, how can you get started achieving the same results?

This guide will walk you through ideas on how to integrate AI into your PR, Content Marketing, SEO, and Digital PR strategy.

Artificial Intelligence In Content Marketing:

How can AI technology help writers?

Artificial intelligence can be used in marketing to predict consumer choice and create personalized content based on data.

AI helps writers:

  1. Assess topical authority with competitive intelligence. How hard is it to rank for specific topics based on the content you have previously published? If you have never written about a topic, it will be harder to rank for it on SERPS. Keyword difficulty is a generalized metric. Artificial Intelligence (AI) helps fill the gaps and aids users in understanding how topical authority impacts rankings.
  2. Content planning: What content should I be writing in the first place? The only way to understand the topical gaps in your content is to analyze a site inventory of every page on your website. AI and GPT-3 analyze thousands of documents and data to make these recommendations.
  3. Historical Momentum. AI content tools help you understand what you should be writing because you have historical momentum in SERPS. After reading the data, you should be able to say, “Here is what I should do to have predictable outcomes.”
  4. Optimize content briefs. AI will help a content marketer to understand: ‘When I execute on this piece of content, it will be successful because of historical momentum and data analysis. This process is achieved through AI content briefs, which results in increased content confidence for predictable B2B content marketing outcomes.’
  5. Accelerate the process of content optimization. With AI and NLG, writers can quickly know which pages to update the cadence on. Before AI copywriting assistants, none of this was possible without extensive manual labor that robots can now assist with.
  6. Movement + acceleration on SERPs. What ideas and topics should be amplified? AI and NLG assist with the process of knowing what to optimize and how to optimize it. But AI can only take you so far. For the best output, it is critical to have a strategic understanding of the desired input. This is critical in understanding the terms you want to rank for on search engines.

How to use AI tools in Content Creation

Creating a content strategy involves human analysis.

Investing time to optimize content with AI tools drives more relevant leads more likely to turn into larger deals with a higher conversion rate. Why? Because you are creating content based on data and metrics instead of guessing what works.

The quality of execution matters.

How well is the data interpreted and acted on to optimize for quick wins that lead to organic traffic, keywords, and position changes?

SEO and content are saturated.

AI will help optimize content, but ultimately, humans will give you the competitive advantage of information gain required to compete in a saturated market. Ideally, you need both AI and humans working together to have the best chance of winning in a digital market.

How to use GPT-3 and AI tools to accelerate new content generation opportunities to win on Google.

AI content tools help marketers understand how every piece of content relates to each other in a digital ecosystem.

As Andy Crestodina of Orbit Media says, websites don’t rank; pages do.

Understanding how every piece of content relates to one another is critical. It’s not just the individual piece of content but how they all work together. This is where AI can give digital marketers a competitive advantage in machine learning and data analysis.

Skip the Waitlist: OpenAI released GPT-3 to the public this week.

AI Marketing 

What is AI Marketing?

AI marketing refers to the process of using artificial intelligence applications to make informed, automated decisions that are based on data analysis.

What are some examples of how AI could be used in marketing?

AI marketing tools can be used to optimize direct response digital marketing campaigns. 

In the past, when marketers wanted to get the word out about products and services, they relied on traditional marketing channels and tactics including advertising, sales, and broadcast.

Today, Google is the new digital storefront and you can reach customers in a highly targeted way. AI tools can synthesize large amounts of data with automation tools to reach more people in less time with greater results.

GPT-3 AI copy

What is AI-powered copy?

AI-powered marketing copy uses machine learning to generate output based on the keywords and data you put in the input section to feed the machine. Output is dependent on the data the model has been trained on.

The Ethics of AI-Generated Content

AI content optimization tools can help eradicate low-quality content on Google by unlocking strategic insights to uncover topical gaps. What does this mean?

No, the content will not magically disappear from Google. Eradicating low-quality content means dissuading writers from creating more low-quality content by showing them what constitutes low quality vs. high quality in the eyes of Google.

Ultimately, humans must make the decision to stop pumping out low-quality content, which means using AI responsibly and investing in thoughtful content marketing initiatives that lead to information gain instead of quick wins to game the system.

AI-generated long-form content is used to create articles covering an entire cluster of semantically related topics and entities. This level of comprehensiveness is critical for qualifying content to rank in search engines.

Hitting publish and hoping to rank on search engines is not a data-driven B2B content marketing strategy. AI changes the paradigm by providing data-driven insights to create more predictable wins. Content briefs and AI optimization tools can also assist with this process. For example, AI copy tools can suggest additional angles and search intent to cover a topic more comprehensively.

AI can also propose a structure of an article through the creation of AI-content briefs to rank on search engines.

AI writing tools help you understand where your business content is strong in addition to topical gaps in authority and search engine ranking opportunities.  Artificial Intelligence marketing tools help you determine whether you can win on SERPs from a ranking perspective. These tools also show you what you need to do to improve the likelihood of ranking with heat maps, clusters, topical mapping, and AI optimization suggestions to improve your business and healthcare content. 

However, winning on SERPs doesn’t always mean creating more content. It can also mean creating offline authority using traditional PR tactics, too. Writing more articles on a topic is not always enough to rank. 

Expertise, authority, and trust are important components of search rankings. 

Google needs to see that you are an expert in a specific vertical, particularly the one that you are publishing content as a thought leader and subject matter expert in. 

A Public Relations campaign is a critical component to achieving topical gaps of authority. This is a primary example of where strategic digital PR paired with data insights from machine learning can give your business a strategic advantage with competitive intelligence. 

The ROI of Content Marketing + AI tools

AI copywriting tools can also tell you the precise investment you will need to make in content. 

Some publicists and content strategists make the mistake of using keyword analysis tools after writing an article instead of before. While it can be easy to jump into writing an article without writing a content brief or outline first, be sure not to make this mistake.

Do not waste time and money writing content that brings unqualified traffic. Ranking is not the end goal; generating qualified leads from the content you write is. The investment in content marketing is not only the time and resources it takes to write and produce content. It should also include the investment in the tools required to analyze the content after it has been published.

Writing content without having AI tools to analyze and optimize that content is like driving a car without a map. 

AI Marketing Tip: AI content optimization tools should be used before you write a new blog post. 

“Nothing goes from keyword to content. That is the last mile of topic analysis.”-Jeff Coyle

Step 1: Use AI to create a content brief. An AI content brief will guide your writing using natural language processing and machine learning for enhanced results. 

Step 2: Write the article. 

Step 3: Use AI optimization tools after you write a blog post to enhance the content. 

It is important to note that this is an ongoing process and not a one-time job.

Search intent changes over time. Your content needs to change with it to hold ranking positions that correlate to what users are actively searching for. 

Blogging and Artificial Intelligence

What is the difference between AI content writers vs. AI content optimization tools?

AI content writers generate short and long-form content. AI content optimization tools optimize the content you have already written. SEO AI Content optimizations tools propose technical fixes to improve search engine result rankings such as missing H2s, internal link recommendations, and keyword opportunities. AI content tools can also assess topical coverage and propose related topics to cover in your content.

READ: Blogging for Business Tips 

“Personalized difficulty is a metric we have developed to look at all of the content to determine how well your competitive advantage is based on the content you have written around that topic. For example, if you have no content written on public relations and AI on your website, you have a competitive disadvantage to write about that and it will be difficult to rank,” said MarketMuse Co-founder and Chief Strategy Officer Jeff Coyle.

GPT-3, AI and SEO:

How can AI improve SEO?

Effectively utilizing natural language processing tools is much more than AI keyword stuffing. Content tagging and structured data lead to a better foundation for ranking opportunities on search engines.

AI SEO entity-based content optimization platforms are ideal for pairing semantic schema with corresponding machine-readable entities. Google’s NLP API plays a role in many of these tools.

AI SEO tools can assist with:

  • Position monitoring
  • Search intent classification
  • Title tag generation
  • FAQ
  • Product descriptions (e-commerce)
  • Rank tracking
  • Keyword Clustering
  • Marketing Automation
  • Title tag optimization
  • Meta descriptions
  • AI-Powered Structured Data
  • Internal Link Automation
  • Schema Markup Automation
  • Semantic Annotation
  • Content Structure Suggestions
  • Long-tail keyword opportunities

Creating a semantic SEO content marketing plan involves utilizing natural language processing (NLG) AI tools to assess business opportunities. Entity pairing is not only an SEO function; it is a core function of public relations, which is where we come in. We help clients create real-world entity relationships that can serve as a structural framework for SEO entity associations in digital marketing.

AI SEO tools can also find the top search competitors for a topic you want to rank for on search engines.

For example, SEO AI tool WordLift automates your SEO by translating your site into Structured Data, the native language of Google and Bing.

“Annotating content, also known as semantic enrichment or lifting, creates metadata that computers can understand.” Andrea Volpini at BrightonSEO shared how to use GPT-J, GPT-3, or Jurassic-1 to automate SEO.

SEO Automation using GPT3 and Transformer Based Language Models

Will AI replace SEO writers?

No, AI will never replace SEO writers. It can replace human SEOs for the parts of the job that they fundamentally don’t like doing. AI is ideal for the automation of repetitive tasks. But it will not replace the strategic component of their job. The best (and most strategic) output will always come from the human brain.

Does AI Content rank?

AI content can rank when AI tools are used to optimize the content you have written. Content written by GPT-3 and AI tools without human curation can often include incorrect information and thin content. It can also be plagiarized, which would tank rankings.

“Most models’ training data cut off in October 2019, so they may not have knowledge of current events.”

Should I fire my marketing agency because GPT-3 writes better copy?

A large number of Silicon Valley companies are growing on top of GPT-3. AI Natural language generation tools powered by AI claim to change the future of marketing. But does the hype live up to the reality? While AI may change the way marketing teams write copy, it does not change the core fundamentals of a solid content marketing strategy.

TL;DR: Don’t discard your content marketing agency for GPT-3 tools. AI copy tools will help your agency, but they are not a replacement for one. 

GPT-3, AI and PR:

We have always been early adopters of new technology.

From social audio to GPT-3, Ruby Media Group is at the forefront of digital advancements in public relations and social media marketing.

PR firms that understand how to leverage GPT-3 and NLG tools will have a significant advantage over the ones that don’t. It is no longer enough to run a PR firm and shy away from technological advancements.

To have a competitive advantage, you must understand the core principles of SEO, Content Marketing, and PR and how all of these areas impact the other.

GPT-3 is no exception to this rule. It can serve as a secret weapon for Publicists who invest the time and training to understand how to leverage it.

Publicists who fail to understand how to integrate AI and ML into publicity campaigns will fundamentally be at a disadvantage. A tech stack that consists of email, Safari, and Cision simply will not cut it as we enter into Web 3 and the metaverse.

WATCH: SEO and Digital PR Masterclass

How can Publicists use AI?

Ruby Media Group leverages AI-powered PR tools.

The PR agency of the future will use GPT-3 for:

  • Pitch angles 
  • Press release titles 
  • PR event ideas 
  • Product Announcements 
  • Talking points/bullets for media interviews 

New AI PR tools aggregate media queries from journalists and incorporate GPT-3 to reply to queries. Essentially, GPT-3 is used to reply to a pitch from a reporter. We believe that experts should write replies to queries instead of machines. While GPT-3 is useful for idea generation, it should not be used to submit expert quotes to media outlets.

This gets into the ethical question regarding the usage of AI in Public Relations. I am of the belief that subject matter experts should write replies to media queries instead of machines. 

While GPT-3 is useful for idea generation, it should not be used to submit expert quotes to media outlets. 

Reporters are looking for quotes from people; they are not looking for regurgitated content and article spinning tactics disguised as PR. 

GPT-3 has the ability to create a rampant amount of fake news and misinformation if it is not being used responsibly in Public Relations. Just because it can be used this way does not mean that it should be. 

When you assign a quote to a subject matter expert that carries legal weight and liability. A PR pitch generated by GPT-3 can land a source in hot water. GPT-3 does not always provide accurate information. You can potentially be feeding a media outlet fake information tied to a real person. That is a recipe for disaster. 

TL;DR: The use of GPT-3 in Public Relations is a controversial topic. 

READ: 20 Ways AI Could Transform PR 

GPT-3, AI and Social Media Marketing:

Our PR agency uses GPT-3 and AI marketing tools to create:

  • Bios for social media
  • Personal branding landing pages
  • Event descriptions
  • Updated meta tags

Artificial Intelligence can enhance your public relations and content marketing strategy. AI is not a replacement for your agency. It can optimize the work they are already doing for you.

You do not have to choose between one or the other; ideally, you should choose both, preferably an agency with core AI marketing skills. Most agencies work on monthly retainers.

An agency that leverages AI and GPT-3 tools will get more done in less time. This not only saves them time, it saves you money by shaving off repetitive tasks in favor of strategic work. If your social media agency is not utilizing these tools, you are most likely paying for humans to spend double the amount of time on tactics that AI could do for them.

When reviewing your marketing agency, be sure to ask them about their capabilities in GPT-3, NLG, and AI. You do not want to pay for their learning curve. When you work with Ruby Media Group, you pay for results, not for us to get up to speed. We train on our own time, not on yours. If you are sick of stale, outdated, and traditional approaches with lackluster results, contact us today to learn more about optimizing your digital PR strategy with artificial intelligence.

READ: How to Design an AI Marketing Strategy 

Create short-form content at scale

The secret to effectively utilizing GPT-3 is to use it for the ideas vs. as the direct output. Where most marketers go wrong is that they rely too heavily on data that can often be skewed, misleading, or outright wrong because it is trained on an old dataset. Marketers also run into the risk of generating content that can be plagiarized from GPT-3 tools, leading to the question, who is liable for plagiarized content? The agency? The GPT-3 tool?

As a legal-first PR agency, we believe it is best to air on the side of caution when it comes to GPT-3.

GPT-3 PR Tip: Use GPT-3 to develop new ideas and then modify the output.

Need help drafting long-form content? Contact us today

Best GPT-3 AI writing tools | AI Copywriting Software 

What is the best artificial intelligence software?

The best GPT-3 AI writing tool is the one you use. Ruby Media Group has invested heavily in machine learning and GPT-3 technology.

Our tech stack of AI Copywriting tools for content creation and optimization includes:

AI Marketing Writing Tools for Business

Ruby Media Group enterprise clients gain access to our innovative marketing tools. Our agency invests in GPT-3 to bring clients the best marketing tools for marketing automation and AI.

GPT-3, AI and Content Marketing:

Can AI create content?

Yes, AI can create content, but the better use case is utilizing AI to optimize content that has already been created. The reason for this is that some of the content that GPT-3 generates can often include incorrect information, outdated statistics, and duplicate content from other sites.

The content that AI copywriting tools creates may not lead to true information gain. Where AI succeeds is that it can tell a writer all of the semantically related topics to include to add to the topical depth and authority of a piece of content. That level of research would take a writer hours to do in addition to the writing. This is why AI optimization tools are preferable to using AI copywriting tools.

AI content requires heavy fact-checking to ensure accuracy. Although much of the content is great, it requires more training before it can be relied on without the assistance of human writers and editors.

The more human input you provide, the better results you will get from AI copywriting tools.

AI CONTENT OPTIMIZATION TOOLS: GPT-3 AI Writing Tools: 

How will GPT-3 Affect Content Marketing?

GPT-3 will impact content marketing by giving users the ability to create optimized content at scale in a shorter period of time. AI makes it possible to have more predictable content wins.

AI content writing tools assist marketers with creating superior output quality. But you can’t have quality output without quality input. That starts with knowing what you are writing about with a content strategy devised by humans instead of machines. AI-powered copywriting assistants make your writing better.

People who use AI to write content for them without their own thought leadership or industry analysis will find that the content will be regurgitated and will not contain true information gain. An AI copywriting assistant is just that- an assistant, not a replacement.

AI can be tremendously useful in corporate communications when creating marketing copy and revising your current marketing communication plan. But the key takeaway is that the best output requires the best input. The best input is from humans, not machines.

READ: This Company Tapped AI for Its Website—and Landed in Court

AI & GPT-3 in Content Creation

How AI will revolutionize your B2B content marketing strategy

Are you prepared for the future of business? Marketing has changed. It is time to change how you do business to keep up with technological advancements. Social media that does not integrate Web 3 looks stale and dated.

We are constantly thinking ahead and trying new tools and tactics. Other PR firms are focused on the establishment and traditional channels. We leverage new media channels to give you the cutting edge of thought leadership. The old establishment is dead. Let us help update your PR strategy and bring you into the future.

New Trends in AI Marketing & Machine Learning (ML)

How AI Will Change The MarTech Landscape & Shape Agency Growth

To further explore how marketing and PR agencies can leverage GPT-3 technology, I interviewed Aki Balogh, MarketMuse’ Co-Founder. Aki is an AI expert who has been in machine learning for 15+ years with a background in statistical modeling, data analysis, and SaaS.

MarketMuse is an AI-driven platform that helps you build content strategies, find gaps in your topics, and accelerate research and publication.

During our interview, we discussed:

  • Artificial Intelligence (AI) Trends
  • GPT-3 & Natural Language Processing (NLG)
  • How to implement AI into your B2B content marketing plan

GPT-3 AI Marketing

The future of the marketing technology landscape. 

What is GPT-3?

Aki Balogh: GPT-3 or Third generation Generative Pre-trained Transformer, is a neural network machine learning model trained using internet data to generate output based on a data set of 175 billion parameters.

How GPT-3 works

Aki Balogh: At MarketMuse, we built our own machine learning engine that generates text and is not dependent on GPT-3.

Any marketer who uses GPT-3 soon learns that it’s really good for a couple of things and it’s really bad for a whole bunch of things.

It’s not because the technology isn’t promising; it’s just that the technology is not there yet.

Everybody wants to automate their work and we’re all excited about AI, but for practical applications, when your job depends on driving leads and getting a pipeline full, it is important to be clear on what you get (and don’t get) when you use GPT-3.

What should people worry about when it comes to GPT-3 and AI?

Aki Balogh: Anyone can sign up on Open AI’s website, get an API key or a license, and get a hundred thousand credits to generate text and play around with it.

Essentially, you are getting a machine learning system that is great at recognizing patterns. However, it doesn’t understand what those words or concepts are and it doesn’t have any kind of knowledge.

Artificial Intelligence and Machine Learning 

Aki Balogh: For example, when you think about a baby learning about different concepts and the baby walks up to a dog and there are dogs. My pet is friendly and all these associations just form in your brain. When you see dog and pet and friendly and to pet a dog is different than having a dog as a pet.

There are all kinds of interpretive mechanisms in your brain that let you navigate the topics when you see them in written form. When you’re reading an article, she pets the dog and you get a sense for it. GPT-3 often misses a lot of this.

AI & Co-Occurrence

Aki Balogh: GPT-3 or any AI system doesn’t understand any of that context. It has never seen a dog. It can’t even see in the way we can. It doesn’t have the same senses. It doesn’t have any emotions. All it knows is that when it sees the words dog and pet, they tend to come up together a lot. That is a simplification of what the systems do. But years ago, that is where this started with looking at co-occurrence.

If you read a million pages on the web on dogs and those pages tend to have the word pet in them, then there is some commonality or correlation between dog and pet because they co-occur a lot.

That is a very early implementation of Google, but that is where these things started and then data builds on it. GPT-3 spins through a lot of texts including billions of articles. The last engine that is publicly available is 80 terabytes or 175-billion vectors of content. But now the next one (Megatron-Turing NLG) analyzes 530 billion vectors or tokens so there are some limitations, but basically, the machine reads an insane amount of text and makes inferences, but it doesn’t understand what those things are.

The system is really good at drawing analogies and connections between things, but those connections don’t necessarily make any sense because it’s not a real-world situation. It just looks like it.

What is GPT-3 used for in marketing?

Aki Balogh: GPT-3 is great for short things.

  • Titles
  • Short sentences
  • Short stories
  • Product descriptions
  • Email subject lines

It’s great at those things because the data it needs to generate is very small and it doesn’t have to be comprehensive.

The shorter the data, the better the output is going to be.

If you imagine taking all of the data that Alexa and Google Home records and generating a generative model on that, it would have a different flavor to it than models generated on text alone. In the conversation, you have inflection and tone and there is additional data.

READ: A Content Marketer’s Guide to Natural Language Generation

AI-Generated Content: Pros and Cons

GPT-3 Challenges & Limitations

Aki Balogh: One of the issues with GPT-3 is that the things it generates it has no understanding of.  It could totally be wrong. It could miss out on critical information. It’s not so much that it would tell you something that you see and it’s incorrect because you change it, but it leaves out a lot of stuff that is actually correct. That is the most challenging part of GPT-3.

If you have ever tried to build a Top 10 list or The Best ABC in an area, it doesn’t know how to do that. It’s just trying to find cross-connecting pieces.

GPT-3 has not been widely adopted by the public relations industry. Why do you think that is?

Aki Balogh: PR is driven by short, concise messages. So, a quick little soundbite that just sticks in your brain, and it’s very much driven by popularity and brand recognition. That is the opposite of what computers are good at. Computers are good at reading a billion documents and identifying a thousand things that correlate to another 5,000 things. Computers are great at connecting the dots. But the point of PR is a short, pithy phrase that will unlock some idea in your brain and that’s the thing that computers are terrible at.

Trends are powerful. I don’t mind the hype around GPT-3. Is it going to fail expectations? Yeah, of course, but that’s okay. It’s going to advance science and one day the science will be great. One day GPT-7 will actually be great.

Every day a new GPT-3 tool is launched. Is the engine different, or is it just great marketing?

How do people keep creating new SaaS tools with GPT-3 and AI so quickly? Are they modifying the engine?

Aki Balogh: No, they are not modifying anything. For example, imagine there is a Bitcoin called Aki coin that I created in three seconds. I just created an Aki coin and I put it on a thing. Now I’m the only person with the Aki coin and it’s worthless, but it took me 30 seconds to fill out a form. Then it took another 30 seconds for the system to create it and now it’s there and it’s never going to go anywhere.

It’s not hard to build stuff when you don’t actually build it. It’s hard to build something when you actually sit down and create something new like new science.

In everything we build, we try to inject new science and some new functionality that is customized to what the user needs.

When you see ten AI copywriting tools, they all say that they do different things and look different. Are you saying that outside of marketing, the backend structure is very similar through Open-AI?

Aki Balogh: The backend structure is exactly the same because Open AI is the only place to get the data. We have our own, but we don’t have APIs to it and probably other companies have built their own models, but there’s only one public-facing engine.

It’s like going to Google or Bing. You only have two options.

When people buy ten different GPT-3 AI copy tools, are they buying the same tools dressed up differently?

Aki Balogh: Yes, they are buying exactly the same data.

Is this the greatest marketing ploy in all of history then? Because we all keep buying them. Are we being sold the dream that each one is different?

Aki Balogh: It’s all marketing. A thousand percent. People are buying the dream.

Everyone is getting their data from the same exact place and then they slice and dice the data differently. There is only one place to get the data from. You can just go to GPT-3 Open AI, sign up there, and do exactly the same thing inside their environment.

I’m curious what these other tools claim to add to that because they can’t possibly have built anything since they haven’t built the data.

Investing in AI | AI Market Predictions

Silicon Valley has invested millions of dollars in tools that get acquired. Will we see a consolidation of GPT-3 AI copywriting tools?

Aki Balogh: Eventually. If somebody raises real money, then they can actually build their own data, which at that point is different. I was more referring to the startups where it just comes out of the woodwork on a weekend and bam it’s there.

Those teams have not had any time or resources to build anything new, but companies like Copy AI who raised 13.9 million dollars are raising funds to build something that is new and proprietary.

There are probably more engines out there than just GPT-3 and ours, we just don’t have a map of it.

It’s like a race.

Different race cars have different engines. Whichever engine is better will win out.

I would think those businesses are going to be challenging to build because GPT-3 is so big and open.

Open AI has so many contributors and resources that we’re probably going to build stuff faster and better, but, I could be wrong. Maybe there are different applications that they don’t cover that Copy AI would do really well. That’s what these investors are going in.

Having said that, I’ve seen a lot of AI companies that have raised 8 million, 500 million and they don’t return that investment. A lot of them collapse. We’ll find out in five or ten years, but if I was a venture capitalist, I would really scrutinize where I would put my money because just last night I was talking to an entrepreneur. The company raised like 80 million Augustus AI out of Germany.

The whole thing blew up and became a big scandal. The CEO had to leave and he is living in Belize hiding from the German government.

When it comes to hype, hype can be a great tool for consumers if the product actually works and it can be a terrible thing if the product fails.

If you remember Vibram shoes where you’re running and a bunch of people, myself included got injured from it because it didn’t actually work, but the hype was there, but the product was not good.

By the time people realize it, they’re like, oh my God, what have we done to ourselves. And then the product goes away but your injury remains.

What should investors in AI be looking for?  What is sustainable versus what should they avoid? How does one determine what artificial intelligence companies to invest in?

Aki Balogh: Having been an AI investor, you look at the tech and you go really deep and you try to see- is this real?

If you don’t do this level of research and analysis, you’re going to be like Theranos and it’s all a big con and you could potentially take on more risk. If you look into it very deeply, you can find some good places to invest, but the investors have to be technical and go really deep and actually look at the science and think through both the business model and the science.

AI is a new science and the investment in GPT-3 leads to funding for a new branch of science. Usually, science never gets enough funding so that’s why I like the hype. The hype is good because it gets some amount of dollars to actual projects, but then anybody who’s trying to make it up, those dollars are going to get wasted.

The consumers who invest their money and time into it are going to be disappointed. So it’s a mixed bag.

You mentioned if a company receives the funding it will build its own engine. Are you saying that they would be less reliant on GPT-3?

Aki Balogh: They might not even use it at all. At that point, if somebody raises a couple of million to do something, they can build their own engine. I’ll give you an example.

Our corpus, the last engine that we did, trained for two weeks on 80 gigabytes of data on two TPUs and two terabytes of data. Our NLG product is constantly training on new data. It was a certain specification that was good enough for us to generate on, but if we put more money against it, that would cost a hundred grand or a couple of hundred grand for anyone to do.

But if we put in ten million against it or a hundred million, we’d get a much more workable system because our purpose is more customized to what our customers need versus a general purpose.

We don’t have to be Google or GPT-3; we just have to work really well for the specific purpose of taking an existing content brief and expanding it through texts. There’s a lot of stuff in there.

We built the NLG engine around our product, but what GPT-3 is able to do is it can write code. It can make limericks, it can make songs or create images.

It’s a more general purpose. They are doing basic research and general science that benefits everyone. We’re doing applied research and applying a branch of science to solve a specific problem for our business users. And that’s what these other startups would do.

If it works, then it’s great. I think we will see a lot more GPT-3 companies in the future. They just won’t be GPT-3 necessarily because they will have their own thing.

Build Your Own AI Model | Artificial Intelligence in Business 

Aki Balogh: The future is getting away from GPT-3 and building your own model.

Take what is there and build on top of it. If anybody just copies exactly what is there, then it’s not going to be differentiated if they take what is there and build it and customize it and build other stuff so that it more accurately solves the problem their users have, then it’s going to be even better.

But, if anybody just ripped something off and put a marketing site around it, I don’t have a lot of faith in that.

SOCIAL AUDIO, AI & GPT-3

What role can GPT-3 and AI play in Social Audio?

How to use AI to optimize content from social audio platforms

Aki Balogh: There are two things machines can do.

They can either ingest things and analyze it and find patterns or they can generate things based on patterns.

In the domain of social audio, you could have a GPT-3 audio podcast device.

In order for any kind of AI to really make an impact on social audio, it would have to have access to all of the recordings.

If somebody had access to all of the recordings and data, we could analyze it. But we don’t, so we can’t.

How to take social audio conversations and build a different GPT-3 engine based on the data. 

AI in Social Audio processing: The Future

A GPT-3 engine could be trained on all of the conversations that are recorded and transcribed.

Step 1: Record conversations. Record all of your conversations on social audio platforms.

Step 2: Turn the data into a text transcript and review the data. Parse through that data of all the audio recordings. Turn the audio into text because GPT-3 is done through a text system. Google does this for free on Google Translate.

Step 3: Train a GPT-3 model on the data. Load all the transcripts and train a GPT-3 model on that data. This would have a different flavor to it from a GPT-3 model that is trained on news, Wikipedia or web documents. It would essentially be a different kind of GPT-3 engine.

GPT-3 Social Audio Tip: Plug in room titles into an AI engine to generate more ideas for future rooms.

Over time, you build up a lot of content around your core area of subject matter expertise.

It’s hard to get someone to listen to an entire show. If you don’t like listening to podcasts, you’re not going to listen to it, but you may listen to a highlight or an audiogram and you can stitch those together and make an interesting and interactive piece of content.

READ: How to Create a Social Audio Digital Content Strategy

CONTENT ROI

What is the best B2B marketing strategy for quick wins: SEO, Content Marketing or PR?

Every client questions the ROI of content and every dollar spent on marketing. Whether it is PR, content, SaaS tools or a paid PPC campaign, people want to know how to make the most from their marketing investment. How do you explain the investment in content when many business owners think they will never rank for anything in such an oversaturated market?

Aki Balogh: If you have a specific niche, if your business is as good at something specific and it’s better than other businesses, you have a way to win on Google and in ads if there’s a basic commercial insight that makes your business viable in some way, then promoting that insight and writing it down is definitely a good move.

If your business copies what somebody else does and they’re stronger and more well-funded, then yeah, don’t waste your time. But if you have something new, you can write content and it will add something to the Internet and then people searching for that specific topic or topics related to that thing are going to find you.

Yes, there is totally a way to win even if you’re a small business. I know this because we do it all the time. If you search for ‘AI content optimization’, we come up because we’ve invested in that as part of our content marketing strategy.

You have to make an investment and give it some time, but there should be some results if that person is doing that job well and with expertise.

MARKETING ADVICE FOR SAAS FOUNDERS

Aki Balogh: Where a lot of business owners get tripped up is to do marketing, you have to waste some money.  Half of the money you invest will turn into new business and the other half may not perform. People are uncomfortable wasting millions of dollars but you won’t know what performs until you figure out what doesn’t perform.

You can’t see what turns into sales until you figure out what will not turn into sales. You won’t know what a low-yield return is until you find high-yield opportunities. But that’s just life. Instead of focusing on what didn’t return the ROI you wanted, focus on the marketing activities that did, and invest more in that area.

The process of figuring out what performs shouldn’t mean that you shouldn’t try. Instead, you should try on a small scale if you are uncomfortable with the heavy risk that comes with deploying dollars towards marketing activities.

Instead of trying to rank for fifty terms or keywords, try one specific keyword or topic and start there and build up a site around that one thing.

But one is better than zero. It’s a completely different dynamic to do something versus to absolutely do nothing and say, no, I’m not going to do that.

Know what you want to rank for.

For example, MarketMuse wants to rank for:

  • AI content optimization
  • AI content strategy

The MarketMuse blog has 250+ articles pertaining to these topics. “If we rank for other broad terms, it could lead to confusion in our funnel. Good marketing puts people in the right funnel.

Use AI Optimization tools to help cover every aspect of search intent.

There are so many ways to market and it’s very confusing and people have to try at least three or four things to see what works.

Try a couple of things and then one channel is probably going to work out better and then focus on that. But doing nothing is not going to give a resolve.

With the ROI question on content marketing and how can I avoid wasting money and time on this? The answer is you can’t.

The only way to lean into something is by doing it. In the beginning, you’re going to be very inefficient and then after doing it more, you’re going to get better and better. People can do this for free with no budget. It’s just the cost of their time.  We have really dropped the cost that far, but they still need to invest hours into writing an article.

It’s an investment of time, which is also money.

What about people who refuse to invest that time and believe that AI copy tools should do it for them?

Aki Balogh: They will end up still doing it, but now they’re trying to do it more effectively. The AI isn’t human. It is always going to need handholding. It’s an investment of time.

People are hoping for the dream of zero-time investment. Even if AI was reading your mind, it would just give you ideas and you still have to make decisions, then you stop to invest time into it.

There is a time investment, but it is also about greasing the wheels on that time investment.

What do you say to those people that want to hire people to do something and don’t want to put in the time to work with them?

Aki Balogh: I now get the overall theme of where you’re coming from.

In this definition, AI and agencies are exactly the same things. Hiring something else, a product, whether it’s a group of humans or a system to do something for us, it’s the same exact thing.

AI agency, GPT-3 is the same exact thing. It’s people hiring somebody else.

Maybe they don’t want to think about it because they feel like “I don’t have anything valuable to contribute,” or “I’m not good at writing,” or “I’m not good at PR,” and that’s the challenge of an agency.

The only way a business is going to be successful and make money and make more money is if:

  1. They develop some business model that is hinged on some insight that makes people’s lives better so they want to pay money to buy it.
  2. Figuring out a commercial insight to communicate that value to the broader world so that people become aware of it and try out the thing.

Any business has to do that and that can’t be farmed out. The founders and the people in the business and the shareholders must be involved. Those two things have to exist and no one can figure that out.

If within a business the business owners or employees don’t know what that thing is, then the agency of record is not going to be able to figure it out because that’s why you’re in business to do something better or differently.

But everyone who has a business that has made at least $1 has something unique.

If you have made at least a dollar, you have something unique. You need to put that thing out there because chances are there will be more people who have that pain point.

It’s just that hesitation of, I’m already doing accounting and I’m doing legal and sales. And now I have to do this marketing stuff too, and PR. Put it aside, subjugate your ego and give it a shot.

Nobody is going to figure that out for you. You have to be involved at least a little bit.

That said, agencies can really help frame and guide you down the path. But at the end of the day, that comes from inside you as the subject matter expert about your own business.

What makes for a successful agency-client relationship?

Aki Balogh: When I started my career, I worked at a management consulting firm and that was our differentiation over at McKinsey & Company, BCG, Bain, the big three that have a big brand and all the money in the world. We would sit with the client and work side by side with you.

We wouldn’t just go away and build a strategy and come back and drop a bunch of slides on you, walk away and give you a bill for $500,000. We’d actually sit with you and build a strategy and it would still have things in there that you may not want to do or know how to do, but then we will do what we call mobilization, which means building an implementation map and actual project management program management organization and turn those strategic areas into capabilities.

We would do that in the context of your business, not some theoretical business in a Harvard case study.

Sitting side by side with the client and really understanding what their opportunities and limitations are is crucial for success.

If you’re there with the troops and you’re fighting the battle, it’s going to be received much better and it’s going to be much more actionable and accurate then, let me just run my general ad campaign that I use for marketing companies. Bam, here you go, and here’s a bill.

Companies may hire an agency or deploy AI to not have to think about something, but the most successful project is if they pull that service provider in or that technology and think deeply and invest the time to build that relationship and train that person on their business and then they learn and open their minds and learn new skills from their agency.

The combination of the two is going to make for the absolute best quality output. And that’s not a waste of money or time.

It’s the dream of not having to work because a machine does it for you. People are going to go to great lengths to live that dream. Even if it doesn’t exist today. It’s that feeling of, wouldn’t it be great if a robot just did everything for you?

The closest we have is Amazon. You click something, you get it within two hours. It’s amazing. But there are a lot of humans involved in the process, and we don’t see those humans.

The same is true when it comes to agencies and the adoption of AI tools.

 KEY TAKEAWAYS:

  • Human curation is still required for GPT-3 output.
  • AI tools free up your time to create a content strategy.
  • Use your brain to create a content strategy. Use AI tools to assist in the execution of the strategy.
  • The goal is not to create more content, faster. The end goal is to create more in-depth content that leads to true information gain.
  • AI copy tools are not a replacement for journalism. Do not rely on AI copywriting tools for ethical considerations when reporting. Use your brain and the traditional model you were trained on as AI can often make mistakes.
  • Automatically generated AI content that is intended to manipulate search engines instead of help users may run the risk of being penalized.

ABOUT RUBY MEDIA GROUP | CONTENT MARKETING AGENCY

AI-powered content creation services

How much business are we losing by not being higher up on the SEO food chain on Google? This is a question every business owner should be asking themselves. 

Ruby Media Group uses AI tools to amplify companies that aspire to have the level of celebrity required to do business without the constraints of traditional marketing channels. Maybe you’re not currently a celebrity in your niche. How do you find the right people in the niche to help you become a go-to thought leader? Ruby Media Group will raise your profile utilizing content marketing, digital PR, and social audio strategies. We build high-quality content that converts. Contact us today for a consultation on your 2022 B2B Content Strategy.

RMG clients gain exclusive access to our MarTech stack of AI marketing tools. Apply to become a new client today. 

 

Create Content For My Business

 

#

 

FINAL THOUGHTS:

This article was written by a human and not a machine. While AI can help find keyword optimization opportunities, it is not a replacement for real-world boots-on-the-ground learning. To fully understand the language models and tools presented in this article, I immersed myself and absorbed as much information as possible from the AI & NLG experts interviewed above.

This is a primary example of why true information gain will come from humans and not machines. The research for this article included purchasing all of the tools mentioned, testing the tools, podcast interviews, Twitter Spaces interviews, Clubhouse interviews, and months’ worth of research to understand this burgeoning industry.

This article was written with the goal of information gain first, and ranking second. The humans I spoke to throughout this process helped me to piece together all of this information. This is where the role of human entities comes into play.

The best entities will always be people and not machines. Humans are not as fast as machines, but the end result is a well-researched article that hopefully adds true information gain on this topic.

 

 

Statistics:

Verified Market Research: Writing Enhancement Software Market Size And Forecast