The capabilities of generative AI are expanding at a Cambrian rate. Hugging Face, Dall-E, Stable Diffusion, GPT-3, OpenAI, and dozens, if not hundreds, of other projects, startups, and organisations are developing AI engines that literally produce video from text, create art, music, and synthetic humans, upending our ideas of creativity, art, public domain, copyright, and the nature of reality itself.
Glimpse AI CEO Alex Cardinell told me recently on a TechFirst programme that this is just the beginning.
The ultimate product that AI can produce is more AI.
The most thrilling aspect of it all, according to Cardinell, will be when AI reaches the stage where it can begin creating the code that will advance its own intelligence. And that’s where the ultimate singularity is, when it can begin to improve itself more than humans can, when it can kind of set itself to develop itself. It’s tough to imagine what society would actually be like in such a situation given that we’re already making significant progress in that area. But I believe that will happen to most of us at some point in our lives. And as for me, I can’t imagine anything more thrilling than that.
Exciting best describes that.
Another is dreadful.
Of all, the singularity is that hypothetical moment in time when the curving graph of technological advancement takes off like a SpaceX Falcon Heavy rocket and hurtles almost vertically into an unfathomable future of rapidly accelerating invention and change. It is hard to predict whether this will occur and how it will develop if it does, but even if it does, there is no assurance that mankind as we know it would endure or that the enormous AI entities that could result from such an explosion would be helpful to life as we know it.
What we do know is that generative AI is currently exploding in a new “Silicon Valley gold rush,” as venture capitalists pour hundreds of millions of dollars into 150 (and counting) generative AI startups in the fields of writing code like a software engineer, chatting like a company service representative, generating images like a photographer, creating music like an artist, selling like a sales rep, diagnosing disease like a doctor, and (gulp!) even (gulp!) diagnosing cancer.
And innovation is happening faster than ever.
“It evolved from sort of this novelty that you kind of laugh at to something that’s actually generating really high quality stuff that people can utilise, that can even in many cases demonstrate [the] same quality as humans in what was basically like six months,” Cardinell told me. And in terms of what is then feasible in the future, it’s really only the top of the iceberg.
Cardinell’s firm produces a text-creation AI that he has verticalized into two products: WordAI, which rewrites content in a distinctive way in a customer-defined style, and Article Forge, which develops SEO-optimized blog posts on just about any topic based on a provided keyword.
When I used Essay Forge to write an article about Apple’s new iPhone 14 Pro, it included subtle faults that human writers wouldn’t have, and I was grudgingly impressed by this as much as pleased.
Writing that Apple has already issued the new model and that the price is anticipated to be greater than that of earlier versions are two examples.
Even so, the work was solid and could readily serve as the foundation upon which a human writer could add to, modify, revise, or otherwise improve. Providing content that no one will ever write but might find a valuable human audience is another use for technology like this.
Cardinell says that the alternative to consumption is non-consumption.
In the long tail, when there is basically non-consumption with the alternative and you couldn’t afford to generate that content in the first place, is where AI can really be liberating, in my opinion. And you might probably guess that given these incredibly esoteric subjects. Even for news, you could imagine that if just 20 people are interested in reading an article on something that happened in your neighbourhood, a human being writing it would be pointless.
Another option is to create a post or article on-demand about an esoteric subject. Because, as Cardinell notes, Google estimates that 25% of people’s search queries are terms that have never been searched before.
As a result, Cardinell says, “you can see how in many respects the next iteration of that could even be, well, I want something written for me that’s covering a topic that no one else has ever written about.”
The key, as always, is not necessarily what technology is doing right now. The key factor determining how capable it can be tomorrow is the rate of improvement.
According to Cardinell, “a couple of years ago, when you thought AI that could create content, you were imagining this kind of machine-written gibberish.” Even your four-year-old child could have written something more impressive, it appeared.
