In the summer of 2010, I sat down with my son and had a serious conversation with him about what he wanted to become in life. He was at a point in his education where the classes he decided to take in the present would determine what kind of studies he could pursue in the future.
At that moment, as a professional with two decades of experience in the technology space, I had a front row seat to the ways machine learning and artificial intelligence were going to disrupt nearly every industry.
I vividly remember suggesting to my son that he pursue a career in the creative space, to do something where he could work with his hands. I told him that a career in the arts or creative services was most likely to be the final frontier where humankind could outpace AI.
Remember, this was 12 years ago during a time when the conventional wisdom was that AI would automate most repetitive tasks, targeting administrative and middle-management jobs. This projection was based on the great flood metaphor of Hans Moravec (Hans Moravec’s 1998 essay from the Journal of Evolution and Technology) describing the “landscape of human competence” and was later illustrated in Max Tegmark’s book, Life 3.0: Being Human in the Age of Artificial Intelligence.
Boy was I wrong.
Today, AI is making great strides in decision-making, yes, but even faster in the domain of arts, and the timeline of the rising tide is compressing. While art and the practice of creating still requires a higher level of thinking, we can no longer view creative lines of work as solely driven by people. The application of AI in the domain of the arts is spectacular, whether it’s creating something that is “written,” “spoken,” “drawn,” or “composed.”
The proliferation of AI has upended our preconceived notions of what only humans can do and we’re finding more ways to automate tasks.
The first notion of this back in 1997 was when Kasparov, during his chess match against Deep Blue, noted signs of a mind in Deep Blue and he worried that there might be humans behind the machine feeding it information.
Another good example of this theory in practice was when in 2016, the Alfa-Go computer program solved Go, the Chinese board game. One spectator described what they witnessed as “moves that were definitely not human, but beautiful,” while others went so far as to say the moves were touched by the hand of God.
It was the first time we saw a glimpse into the potential use of AI as a creative force. Visual AI especially is making quantum leaps and this rising tide will sink those who choose to not apply it in new and innovative ways.
Tesla is one such example as a pioneer in car manufacturing. The investments it is making in autonomous driving is pushing the boundaries—and consequently other car makers—to leverage visual AI and disrupt the traditional ways of how we view transportation.
New AI tools and models such as DALL-E (or the free version DALL-E Mini), GPT-3, Midjourney, Stable Diffusion, and Playground allow users to create in a wide range of ways; from making a drawing based on a sentence you write as a prompt to generative copy, we now have a whole new frontier to explore.
Auditory creative is impacted as well with AIs that generate soundbites based on a written prompt. AI algorithms visualize media, turning notes and sounds into 1s and 0s, and then find patterns throughout this massive wall of numbers to accurately say a rock song is a rock song—and then extrapolate that data to create new music.
What does this mean for creative services and the arts as it tries to stay afloat amidst a rising tide?
To me, this is an opportunity to embrace AI and other technologies. It can supplement your work like Google is supplementing you every day (Google in the end is an AI). You can use a computer to begin a project and put into vision something you would only otherwise be able to see in your head and then you creatively develop it further. You have to look for new ways to bring AI along and understand where it fits into your business, otherwise you will become like Blockbuster—an industry behemoth that was outpaced by other service providers who came along and adopted new technologies and applied innovative ways of doing business, faster.
In short, AI fundamentally touches everything we do in industry. We’re accumulating data at record speeds and the cost of data storage is so low, we’re able to retain all the data we’re collecting. But we don’t have enough time in the world to go through all of these mountains of data and derive insights that we wouldn’t otherwise see. That is the opportunity AI is bringing us.
So, I have to admit I was wrong 12 years ago. AI will be omnipresent in our platforms, solutions, and the way we create.
Frightening? Well, I don’t know.
For some it will be, for sure. Like the current debate around stable diffusion, people are using the AI tool to mimic real art by typing the artist’s name into the prompter. For example, take Rutkowski—he is known for his epic depictions of dragon warfare and other medieval-like battle scenes. Rutkowski’s name has been used to generate around 93,000 AI images on one image generator, Stable Diffusion. So the question of copyright looms, hence the reason that Shutterstock banned all directly AI generated content to figure out a way to compensate the real artist. This is just the first chapter in what is going to be a very long and heated debate.
And where does software development and IoT, specifically, fit into this conversation?
I believe we are witnessing a convergence of traditional art and ways of creating with technology. AI enhanced coding will become a form of art as painting or creating music with AI, just in a non-traditional sense.
So, what can this form of creative expression achieve?
There are currently start-ups aiming to build whole applications—text to code generative tools—based on the description of a prompt. Others are seeking to develop an auto completion feature on steroids or exploring how descriptions of intended code can be converted into a summary of the function written in natural language
AIs will build your queries, will do your integration mapping (the explosion of APIs is begging for auto-mapping features), and contribute to your visual dashboards. Once the AI is trained in something it will be able to contribute, it can validate your work once the AI is trained in testing. The result? Developers will be more powerful and more productive, allowing them to master greater levels of complexity in shorter spans of time.
AIs will also become close companions for platform users, guiding first-time users and advanced users of systems alike. AIs will be invaluable as you look to improve the adoption and conversion rates of your applications, ranging from trials and freemium products to large-scale deployments. If the current trend of product-led driven sales continues, a higher quality AI-buddy that guides you in the adoption of the product could become the make-it or break-it difference for your business.
As a result, IoT platforms must support AIs on all different levels. Generative AI models will be embedded in your user interface as your go-to assistant just as Alexa is in your home. They will help to support in mapping your device data to the platform and in writing the alerting rules. AIs will assist in writing new ML models that can do advanced analytics of the IoT-generated data, allowing for the development of advanced use cases such as golden batch analysis or predictive maintenance in hours, not days. Your IoT platform will be required to have ML Ops capabilities as well so you can manage those ML models through their lifecycles in the cloud, on the edge, or close to—or even in—the device you operate.
That is still the future…lets pedal back a bit. In fact, lets pedal back to the present and my son. He is now a creative consultant and over a rainy weekend we played with DALL-E and other engines. I was impressed at what we were able to make in a few seconds. I asked him if he was afraid that AI would take over his job. He said not at all. He believes AI makes his work easier, accelerating his ability to accomplish projects from weeks to days and improving his interactions with designers that still build the final artwork. He pointed out that the ideas of what to write in a prompt is still ours to be.
As he explained this to me, my favorite quote of Arthur C. Clark came to mind: “Any sufficient advanced technology is indistinguishable from magic.” With that in mind it seems, in the end, we will all be magicians. How nice is that?