Demo this! Everything you ever wanted to know about AI (but were afraid to ask)

Industry NewsCategory
11 min read
Mike Demo

There is so much that people want to ask about artificial intelligence (AI), and with ChatGPT’s meteoric rise, the questions continue to increase. You can’t turn on any network news channel – or even fintech news – without hearing about ChatGPT.

However, before we talk about robots replacing human writers, let’s bring up one profound question: Why do we write?

Why do we write?

Before choosing how we write, we must understand why we write. What’s our end goal? Compare writing to building a website: Why do we want this page? What do we want people to experience from this page?

We write to express our experience, expertise, trustworthiness, and authority. We write to show that we’re the person to hire. We write to educate our audience. We write to convince and to persuade. If all of those terms remind you of Google’s E-E-A-T, then you’re not wrong. We’re talking about findability – otherwise known as SEO.

“You know who should be most worried about ChatGPT and generative AI?  Marketers who lack creativity.”
Spark Toro

If you’re writing to be found and hired, don’t fully rely upon AI writing tools. Maybe you’re done reading at this point, but you may have more questions. AI brings up tactical questions for marketers and philosophical and ethical questions for the rest of us.

Let’s start from the beginning: What is AI?

Related: AI prompts for small business owners

What is AI, and why should I care about it anyway?

AI is a computer program that can learn from data points. Those data points may come just from you – a user logged into a SaaS or local program – or from a collective amount of data available on the web.

So no, your washing machine isn’t AI. Just because it has an app to tell you when to get your clothes out of the dryer doesn’t mean it has the intelligence to collect, sort, wash, dry, fold, and put away your clothes.

Now that I think about it though, maybe that’s something we could all get behind.

AI is part of our lives now, and it’s here to stay. Want to take a couple of seconds to clap to your favorite song while driving – hands-free? No problem if you have Cadillac’s Super Cruise. Tired after a long, relaxing dinner out? Let your Tesla drive you home.

“Some of these systems, like Cadillac’s Super Cruise and Ford’s BlueCruise, offer hands-free driving on pre-mapped highways.”
—Cars.com

Need to check your grammar? Grammarly’s AI system checks that and will even make suggestions to premium members. However, aside from removing plural agreement, spelling, and clarity errors, Grammarly’s AI often removes the quirks and nuance that give writing a personality and a voice. It definitely doesn’t like repeated sentences or alliteration, and that surely makes for some dry reading.

“Grammarly’s AI system combines machine learning with a variety of natural language processing approaches.  Human language has many levels at which it can be analyzed and processed: from characters and individual words through grammatical structures and sentences, even paragraphs or full texts.”
Grammarly

What are the main applications of AI today, and how will it affect our lives tomorrow?

AI is involved in various things such as medical diagnosis, self-driving vehicles (as we’ve mentioned), financial investing, and more. It leaves us asking ourselves, where isn’t there AI?

Coming back to marketers, how will AI affect them? There’s always some program coming for our jobs, mundane or not, and the hubbub about ChatGPT isn’t new. The artist formerly known as Jarvis.ai (Jasper) was popular a few years ago. WordPress has Bertha.ai (a partner of Codeable) specifically to help website owners – and even professional copywriters.

“Your writing will be fuller, have great context, and as the editor, you will be able to write faster, better, and easily increase the ROI for your time and the cost of Bertha Pro.”
—Bertha.ai

Many of us have been using Lumen5 for a while to create videos from our blog posts, and ChatGPT4 will also add video features.

That all sounds like exciting progress allowing us to move from being knowledge workers to design thinkers. The effects of AI are unknown, but again, science fiction has been bringing up ethical dilemmas for years.

Do we allow the machine to do X if Y is a casualty? What if X is only a 70% chance of saving 500 people and Y is the almost certain death of a friend? (Okay, you got me. I’ve been watching Space: 1999 on Amazon lately.)

Does AI really think for itself, or is it just a fancy calculator?

There’s weak AI, and there’s strong AI. Simply put, strong AI refers to what we know from science fiction: namely, a machine that can solve problems for any question asked. It’s still pure fantasy and will probably remain so — for now. Strong AI would be HAL 9000 in 2001: A Space Odyssey.

On the other hand, we encounter weak AI every day. Weak AI is Alexa, Siri, and Google Home. These are algorithms – and that's all AI is, a very complex algorithm – that can answer specific questions whose solution paths they have learned independently beforehand.

AI has no consciousness and shows no understanding (yet). Maybe it shares the latter with strong AIs like the Terminator.

So what distinguishes an AI from a simple program? Typically, programmers write code in the language of their choice, which consists of a set of arbitrarily complex instructions: if this, then that. For example, if the user presses “send,” send the email to server X. These systems are rule-based.

Pocket calculators are relatively simple and have no AI whatsoever. The first pocket calculator came out in 1969, and many programs are available on the internet from C to Python, if you want to program your calculator. A pocket calculator isn’t learning anything when you ask it to add up your receipts – it’s just giving you the results.

In the case of AI, on the other hand, programmers aren’t specifying every single step. They write an algorithm that is capable of creating these steps independently. Thus, we use the term “intelligence”, and we call it artificial because it isn’t organic.

So if AI doesn’t think, how does it solve problems?

AI usually doesn’t write its own code; it just modifies certain parameters within its code to find general patterns in data. That will soon change, however.

“DeepMind entered AlphaCode into online coding competitions.  In contests with at least 5,000 participants, the system outperformed 45.7% of programmers.”
—Science.org

Certain problems are so complex that it’s almost impossible to write the code to solve them by hand. An example is image recognition, used in social media platforms like Facebook.

No programmer in the world could write a set of instructions that always recognizes how I look, whether the photo was taken at night, at the beach, or in a car. In a rule-based system, this outcome would be impossible. After all, the programmer would have to know and describe all possible images of me in advance.

So, we can teach an AI how to recognize me, but not specifically. The AI also doesn't know every picture of me, but it can learn from existing pictures what I look like and then transfer this rule to new pictures in order to recognize me.

What do you mean by “machine learning” and “deep learning?” Are these just marketing buzzwords?

Advances in artificial intelligence are taking off. Only a few months ago we were still talking about NFTs and Web3, and now it's all about ChaptGPT and AI in our SEO. This boils down to two concepts: machine learning and deep learning.

These terms are often thrown together, making them seem interchangeable, but they’re as different as retargeting and remarketing ads.

What is machine learning?

A simple example of machine learning is a video streaming service. To decide which new videos to recommend, the service's algorithms engage in a learning process comparing your viewing preferences to others with similar tastes.

Machine learning utilizes algorithms that analyze data, learn from that analysis, and apply what they learn to make informed decisions. We can apply this to a wide range of automated tasks in most industries: IT security companies searching for malware, personalized marketing, Tesla automated driving, financial investing, and more.

Machine learning algorithms function a lot like virtual personal assistants. It’s okay for what it does, but things get interesting when the computers learn new tricks. In this case, we're talking about deep learning.

“We build our generative models using a technology called deep learning, which leverages large amounts of data to train an AI system to perform a task.”
—OpenAI

What is deep learning?

Deep learning is a sub-field of machine learning that handles algorithms called artificial neural networks modeled on our brain’s functioning.

In machine learning, the programmer intervenes in the data analysis and the decision-making process. In deep learning, we provide the information and document the processes, but the machine is responsible for decision-making.

Deep learning is slightly scary (imagine the Terminator movies) because it’s impossible to trace which decisions were made based on which data because the machine optimizes the decision rules automatically and independently – without us!

This creates a lack of accountability. This generates ethics and legality issues in many industries, including financial investing.

“In addition, some industry participants have expressed concern that AI trading models across the industry may start to learn from each other, potentially leading to collusive activity, herd behavior, or unpredictable results.”
—FINRA

How can I use AI to improve my business operations and save time on mundane tasks?!

You can use AI to improve business operations. That’s its appeal. Nobody wants to pay a robot to vacuum around the house until they realize the time that they would save. AI is the same; it’s a tool.

AI helps transcribe online meetings or videos or create social posts.

AI suits programmatic, repetitive tasks, like buying ads on Google. If you tag your data correctly, AI can fill in the gaps in your first-party data. However, privacy concerns exist.

Can robots ever replace humans, or will they always come second?

This answer is easy. Robots aren’t sentient – or are they? They will never replace humans – fully. Sure, the future of a robot maid we were all promised when watching The Jetsons or a robot lab assistant like in Lost In Space sound great. Science fiction has been tackling the AI versus humans theme forever.

Do you get chills when you read, “Open the pod bay doors, HAL?” I sure do.

If we yield editorial control to these neural networks – artificial intelligence – then what vulnerabilities do we open ourselves up to?

If artificial intelligence learns from the data available, it also learns from the not-so-good parts of humans. This means it’s reading the mean people on Reddit, Twitter, Facebook, and Nextdoor. It’s reading our emails when we complain about our bosses. It’s reading our messages when we call someone names.

“It also results in a much bigger issue that affects all of us.  This content often isn’t diverse and inclusive at all.  It’s created by AIs that were trained with biased content.  And this content is often written by the same type of person.”
—Sam Alderson

When AI learns from people, it also learns about betrayal, lies, and bullying. It learns how to gaslight, ghost, and gamble, just like us.

So, is AI second best to humans? Maybe not. Though they have a carbon footprint the way we do (electricity, space), if they become emotional and sentient, they also take up resources with emotional management (we call this friendship) – and gosh, isn’t that a lot of work for a machine?

Whether robots are tools, friends, or foes, only the future will tell – and the future is what we make it. So, what’s your next move?

Disclaimer: When using AI tools, avoid entering sensitive information and always review output for accuracy. Additionally, this content is for demonstration purposes only; it does not represent any affiliation, endorsement or sponsorship with the companies mentioned. All trademark rights belong to their respective owners. Third-party trademarks are used here for demonstrative and educational purposes only, use does not represent affiliation.

Products Used