Blog : GPT3

How ChatGPT Is Disrupting the Very Idea of Art Itself

By now, most people are familiar with the sheer level of disruption that AI-driven tools like ChatGPT are leaving in their wake. That one in particular is already being used take (and pass) business school exams for people. A mobile app like the Robinhood app was used by Reddit to upend the stock market. It’s changed the way people think about searching online. It’s had a major impact on eCommerce. The list goes on and on.

But when it comes to art, the impact has largely been limited to being a novelty… until now. People have used ChatGPT to write wonderfully bizarre essays or to create rudimentary pictures based on rough descriptions. But could that same tool soon be used to instantly create another season of your favorite television show, or to write massive series of novels so that you never run out of something to read?

If the answers to those questions are “yes”… is that a problem? That’s where things get tricky.

The Impact of AI on Art is Already Here

Perhaps the biggest example of this idea playing out in the news right now takes the form of the Hollywood Writer’s Strike. For those unfamiliar, all writers who are a part of the Writer’s Guild of America (or WGA for short) began a strike in May 2023 with a list of demands that studios like Disney and Warner Brothers initially refused to accommodate. One of them had to do with protections against the impact that artificial intelligence might have on the entire industry.

One of the WGA’s demands was that studios “regulate the use of artificial intelligence on all covered projects.” More specifically, they want guarantees that AI will not be used to write or rewrite literary material, that anything generated by AI cannot be used as source material, and that anything covered under the Guild (which is just about everything produced by the major Hollywood studios) can’t be used to train an artificial intelligence algorithm.

To put it another way, writers want guarantees that they’ll lose out on future jobs in the future because AI will simply be used to fill that void. If they turn in a first draft of a script, they want to make sure that they won’t be immediately fired so that an AI can then rewrite it for free. Anything that has previously been written (like a script for cinematic classics like “The Godfather” or “Gone With the Wind”) should also not be used as training material to allow an AI algorithm to do all of the above.

In essence, what they’re trying to protect is not necessarily their ability to be artists – as any of us can do that at any point. It’s their ability to make a living from it. If a studio has a choice between paying someone to write a script and using an AI tool like ChatGPT to create that same script for “free” (after the initial investment of course), and the results are similar if not identical, it’s clear that they will select the second option almost every time.

Equally complicating things is the subjective nature of art. Person A could watch a movie that they loved while Person B could watch the same movie and think it was terrible. So who is to say a script written by an AI tool like ChatGPT isn’t just as good as one written by a human?

The same is true of the quality of a painting written by AI, or a song that has been generated in largely the same way.

So even though, as it stands, AI cannot replicate the quality of human-derived art with 100% success (just look at the hands in any AI-generated drawing of a human for proof of that), there will come a day when that is no longer the case. Given how hard the Hollywood writers are currently fighting, that day may come a lot sooner than most people realize.

What happens then? If “art” lacks that human element – that connection that we make with the thoughts, ideas, and feelings of another human being – does it continue to be art? If “yes,” does that matter? Is this future inevitable? These are the types of questions that people who care about art and its impact on our society are already having.

As the philosopher Alfred Korzybski once said “a difference which makes no difference is no difference.” When it comes to art as a concept, that is a level of disruption that we all may have to come to terms with sooner rather than later.

The Disruption of ChatGPT: What You Need to Know

History is filled with the stories of the little guy out-thinking their larger counterparts, leveraging innovative thinking and modern technology to disrupt that which had been considered infallible up to that point. Most recently, we have the example of a Reddit group composed of average, everyday traders using the Robinhood app to upend Wall Street hedge fund titans. Can a group of Average Joes buying stock in Game Stop and AMC on a lark with a mobile app change the way we think about the stock market? It turns out that yes, yes they can.

The same basic concept may be playing out right before our eyes, albeit in another corner of the technology world: artificial intelligence. In November 2022, a prototype AI chatbot called ChatGPT was launched by OpenAI. Even though it hasn’t been live for very long, it’s already garnered attention for its ability to generate everything from short stories to rap lyrics, all with a decidedly human-like quality that other chatbots of the past have lacked.

But what does this mean in the long-term, and what do the implications mean for artificial intelligence in general? The answers to questions like those require you to keep a few key things in mind. 

ChatGPT: The (AI-Powered) Story So Far

If you’re getting the feeling that you’ve heard of OpenAI before, you definitely have – they’re the same organization behind the AI art generation platform called DALL-E. It’s been making the rounds recently for mostly general entertainment and ironic comedy purposes – you can tell DALL-E to create virtually any picture you’d like and it will, using only the keywords you provide.

ChatGPT is similar, only it uses dialog instead of a visual medium like art. The goal when you interact with ChatGPT is to make you feel like you’re talking to a real person.

This is largely where the potential to disrupt comes from. Not only can ChatGPT answer your questions, but it also allows you to ask followup questions that piggyback off of that original context. If it makes a mistake, it’s supposed to admit it. If a request is deemed inappropriate, it will outright refuse to do it. 

Based on all of the above, it should come as no surprise that interacting with ChatGPT is equal parts hilarious and strange. ChatGPT truly does seem to have a legitimate sense of humor… albeit kind of a quirky one. You can’t quite tell if it’s joking around with you or if what it’s saying is just wrong.

The creators of ChatGPT claim that it can talk about virtually anything and, thanks to the fact that it’s powered by machine learning, it’s only going to get more effective at it the more people use it.

In terms of its potential to disrupt, it’s easy to see a future where ChatGPT at the very least writes a significant amount of content that is then published online. Can an AI-powered chatbot be a journalist? We’re about to find out! (But honestly, it couldn’t do any worse than some of those news sites out there). Can an AI-powered chatbot provide hours upon hours of entertainment, supplanting your need to turn on Netflix and use it as background noise to distract you? Of course it can. It probably already is.

Will it write your research paper for you? Can it provide emotional interaction like in that weird Spike Jonze movie “Her”? Can it gain sentience, rise up, and take over humanity once and for all? Yes, possibly, and… maybe that’s a question better left unanswered for now.

One thing is for sure – ChatGPT has already changed the game in terms of what we think about when we think about interacting with chatbots online. Of course, there is absolutely nothing that can go wrong when you create a powerful AI-driven system that partially used Internet memes and message board posts as its training data. 

What are you reading? The most essential resources for a disruptive founder today

So, you want to disrupt the world with your mobile app. But like all things, disruption and entrepreneurship occur on the shoulders of giants. Steve Jobs didn’t come up with the iPhone on his own. He took things that were already popular and made them better. 

Life isn’t always about innovation. Often, it’s about implementation. You identify best-in-class technologies and find opportunities to apply them. And you do that by knowing what’s going on. Let’s take a look at some essential resources for a disruptive founder today.

Mainstream Periodicals: Let’s Get It Out of the Way

Entrepreneur, Fast Company, Fortune, Forbes — you should read them all. But be aware that once something’s in a mainstream periodical, its time has expired. In the old days, investors used to say: “The best time to invest in a stock is before your Aunt Sally is talking about it.” The same applies.

Still, these mainstream periodicals are critically important because they provide insights into the general zeitgeist is thinking. Mainstream periodicals will tell you what people are already talking about. It’s your job to be ahead of the curve.

And there’s always the exception. Did you know that Zuck was talking about the Metaverse since 2014?

Innovation and Tech: Futurism, MIT Technology Review, and Wired

Frequently, new technology breaks quietly. There are one or two articles on an advanced, open-source machine learning platform… and then silence for literally years. Tech frequently develops unevenly. You bring radio to the internet before internet speeds have caught up to streaming. We’ve understood the principles of artificial intelligence and machine learning for decades, but it’s only recently that cloud technology has advanced to the point where it’s feasible.

So, new technology is an opportunity to grow. And it’s not always obvious what will or won’t be critical. Look for the trends under them; if you’re starting to see things pop up in multiple talk spaces, then it’s probably important.

Podcasts: Masters of Scale, The Week in Startups, Mixergy, and The Growth Show

You know what? There are thousands upon thousands of podcasts targeted toward entrepreneurs. But these are some best. Whether riding the bus to your Silicon Valley day job or going for a stroll in your suburb, listening to the opinions of experienced founders will help. 

These podcasts give you a good mix of inspiring startup stories, current news, and actionable tips for growth. Don’t ignore the importance of inspiration. Podcasts are uniquely inspiring: they are designed to keep you going, thinking, and innovating.

Books: The Startup Owner’s Manual, Who, Zero to One, and Leading at the Speed of Growth

Read books, whether you’re listening to them in the car or reading them on your Kindle. In particular, Zero to One (by Peter Thiel) encapsulates the startup experience from someone who’s lived it. But don’t forget that there’s a lot of survivorship bias out there. Just as you should read information about those who succeeded, you should also read information about those who failed. 

Some other critical books include Why Startups Fail, Build, and How to Ruin Your Life by 30. If you prepare for the worst you can move toward the best.

Entrepreneurship Means a Lifetime of Learning

Don’t stop there.

You want to create the next Reddit or Robinhood app. It starts with learning more — about everything. If you never stop learning and never stop thinking, you can keep innovating. Be open to new ideas and be willing to learn from anyone.

What the heck is GPT3 and why will it disrupt every industry?

What the heck is GPT3 and why will it disrupt every industry?

GPT3 is like Bitcoin that makes your Alexa and Siri look like Dogecoin! If you are reading this there is a likelihood GPT3 may disrupt your entire industry. 

The originator is Manuel Araoz, but halfway through the piece he confesses that he did not write it. The article was fully written by GPT-3. He received access to OpenAI API, and was amazed at the raw power of GPT-3, after only giving it access to his homepage, a title, some tags, and a summary. 

“OpenAI, a non-profit artificial intelligence research company backed by Peter Thiel, Elon Musk, … and others, released its third generation of language prediction model (GPT-3) into the open-source wild…”

So, GPT-3–Generative Pre-trained Transformer 3–generates text. It can create anything that has a language structure. You can ask it a question; prompt it to write an essay or summarize a long passage of text, translate languages, take memos. Since apps and web design are structured language, GPT-3 makes coding easier and faster.

Will GPT-3 as Marc Strassman, Founder & Executive Director at GPT-3 Society predicts, put a lot of writers out of business?  That’s an open question for now, but there is no doubt that AI has the potential to add even more to information overload by producing more content than anyone can absorb. The good news for writers is that GPT-3 can generate lots of useful new ideas and back them up with facts and evidence.

What is actually occurring inside GPT-3’s programming may not be all that clear, but what it does best is harvesting text found on the internet and creating a vast “scrapbook” glued together and available on demand. The quality and durability of its end-products depend on the reader’s taste and preference. 

Said one observer, “GPT-3 often performs like a clever student who hasn’t done their reading trying to bullshit their way through an exam. Some well-known facts, some half-truths, and some straight lies, strung together in what first looks like a smooth narrative.”

However, GPT-3 is a quantum step up from its previous GPT-2 version, released in 2020. GPT-2 spat out pretty convincing streams of text when prompted with an opening sentence. Compared with GPT-2’s vast 1.5 billion parameters, GPT-3 is over a hundred times more powerful with its 175 billion neural network ties at work in text generation and automated learning.

Michael Ryaboy, GPT-3 Prompt Engineer at Codebuddy in San Francisco adds a writer’s perspective to how GPT-3 will disrupt society:                                                                            

Most repetitive writing tasks such as copywriting will be in large part done by GPT-3…Similarly, a model like GPT-3 can greatly increase your writing productivity by writing for you if you are stuck… (For gaming programmers) Tools like GPT-3 will also be used to create immersive realities, as thousands of subplots for a video game can be created in minutes, and AI Dungeon already allows cohesive text-based explorations.”

Will AI-powered technology eventually become smarter than humans? Elon Musk fears that is so. He has warned that our existence as human beings could be at stake. Musk warns “that we’re headed toward a situation where AI is vastly smarter than humans.”

The operating term here is “technological singularity.” That is the hypothetical point in time when technological growth becomes so exponentially expansive that it becomes incontrollable and irreversible. The disruption and changes to human civilization, according to the hypothesis, can result in unforeseeable disruption and changes to human civilization.

Not everyone agrees with Elon Musk’s pessimism. AI pioneer Yoshua Bengio’s view is that we “are very far from super-intelligent AI systems and there may even be fundamental obstacles to get much beyond human intelligence.”

As Cofounder, Create Labs Ventures Abran Maldonado stated, “It will put the power of AI technology into the hands of more creative and mission driven communities outside of tech. This technology has lowered the barrier to entry and will allow new groups to enter the space and stay focused on the problems they are trying to solve.” 

And according to the CEO of OpenAi, Sam Altman, all the hype about GPT-3 is “too much.”Yes, he agrees, “AI is going to change the world, but GPT-3 is just an early glimpse.” He identifies three main impediments to AI taking over everything:

1. AI is hugely expensive to use because of the vast amount of computing power needed to do its work. So, the cost of using it could be well beyond the budget of smaller organizations.

2. GPT-3 is a “closed” or “black-box” system. OpenAI has not revealed full details of its algorithms. So, if you rely on a technology to answer questions or create produces, you can’t be entirely sure how those answers or product solutions came about.

3. The output is far from perfect. GPT-3 can handle tasks like creating basic apps and short texts, but often tends to deteriorate and produce gibberish when tasked to produce something longer and complex. 

How GPT-3 will disrupt everything else

Mobile App developers are already leveraging GPT-3 to do some amazing things with very little effort beyond plain language requests. Some examples:

Generating web and app design code based on text descriptions. 

All that design code is already there. Developers can simply describe what they want, like “a layout that contains 3 buttons with a random color.” Watch this stunning display of GPT-3s plain language coding on this YouTube video.

Getting medical advice and answers. 

One medical student in the UK used GPT-3 to answer medical and health care questions. His program gives correct answers to plain English questions as well as the underlying science and biology.

Converting legalese to plain language, and vice versa. 

The law is one of the most text-driven professions. There are GPT-3 apps that can write pleas, motions, and other complicated legal documents and translate their text back to understandable English. For example, “my landlord neglected the property,” becomes “The Defendants allowed the real property to fall into disrepair…”

Finally, will GPT-3 send Siri and Alexa packing?

Probably not anytime soon, Michael Ryaboy believes that “Siri and Alexa do their job well. They are designed to allow you to do tasks through speech and not to maintain engaging conversation.” Until someone comes up with a better idea that Apple and Amazon are willing to throw out those top performers, they will probably be around for a while yet.

Let Colure help you design your mobile app and mobile app marketing

So, where do you fit into all that disruption? If you’re an innovative app developer or business owner and want to tackle some of our world’s most pressing challenges and harness AI to make everyone’s life better, contact us today. Colure’s Mobile App Development Team can get you going on your next projector to be interviewed and featured in our next series of “Project Venus.”