Generative artificial intelligence has already been breaking the internet lately with hype about its disruptive potential. OpenAI’s ChatGPT has demonstrated the abilities of AI like never before, and it has got the world suddenly seeing the very real viability of something that was once the realm of science fiction.
By its basic definition of a machine that is able to solve problems and that once involved human input, AI has actually been around for decades already. However, in the past, AI was not able to do things like fully automate warehouses due to the high variability of operations, or write natural-sounding articles due to the high complexity of human language.
When it comes down to it, the missing component to large-scale AI was (and still is) input. As input such as images, text, programs, language models, datasets and more continues to grow, AI will continue to develop.
So far, investors have mainly focused on the obvious beneficiaries of the AI boom, such as Nvidia Corp. (NVDA, Financial) and Microsoft Corp. (MSFT, Financial). However, it is not just the GPU designers and the AI-enabled search engine creators that will benefit from technology that disrupts a wide variety of industries. There are also under-the-radar sectors that will get a boost, including companies that can provide input sources for AI and those that provide education to help re-skill and up-skill workers for rapidly changing industries.
Input sources
Rumor has it that generative AI like ChatGPT will completely replace artists and writers. Quite frankly, I think this is an impossible speculation driven by fear and ignorance. Sure, generative AI can and will continue to perform tasks that previously only humans could perform, but the use-cases will continue to be limited.
If you need a generic image and are not picky about specifics, you can get one from DALL-e or other image-generating AI models that have been trained on the plethora of royalty-free images on the internet. If you need ad copy or another piece of writing where things like factual accuracy and cohesiveness do not matter, or where any minor issues from the AI can be easily and quickly resolved by a human, then ChatGPT can replace some of your writers.
Why cannot AI completely replace artists and writers? Because it needs input, both in terms of training its models and in terms of relevant data. When I ask ChatGPT to write me an article on Microsoft’s latest earnings report, which was published on April 25, it cheerfully tells me “Sure!” and then proceeds to outline the earnings report from the quarter that ended in July 2021. I would have to painstakingly feed it the latest earnings data in order to get an accurate article, and even then it would have inaccuracies and need a human to edit it. With extensive prompts, I can get an AI image in the general vicinity of what I want, but if it hits my exact idea it is a miracle.
That is why I expect there to be a thriving market for input to feed to AI. AI may someday be able to replace the armada of bloggers summarizing and re-stating news articles, but it will not be able to do the same with the writers who are actually reporting the news from information that they and other people gather in the real world. In fact, if generative AI is to someday write us unique human-sounding articles on Microsoft’s latest earnings report, the company behind said AI will need to license the real-time information from news publishers. This could potentially be a positive development for news companies like New York Times (NYT, Financial), as they could benefit from charging high licensing fees to AI companies for news - the earlier you want the news for your AI model, the more expensive the fees.
It is possible the AI images market will always have an accuracy problem, but accuracy can be improved with more input and extensive tagging. Premium AI image generators that charge subscription fees should be able to expand their capabilities much faster than free ones because they can use part of their revenue to license images that cost money rather than just relying on open source. A key beneficiary of this could be Getty Images (GETY, Financial), which boasts 80 million images from contributors around the world.
Online education providers
Another segment of the market that should benefit indirectly from AI is the education sector, especially the online education sector.
Even if people in many countries are required to attend school for a certain number of years, there have always been systemic gaps in education. For example, most younger generations in the U.S. never had the option to attend shop classes, resulting in a shortage of young trade workers in professions such as plumbing, welding, etc., as they just did not realize those careers were even an option.
As technology advances more rapidly, easy and fast access to education on the latest developments is becoming more and more important. Given the disruptive nature of AI, it too will feed into this trend as companies around the world instruct their employees to learn how AI could benefit the company and improve their job performance. It is also important for employees, freelancers and business owners alike to be able to learn the newest in-demand skills for their jobs. One company on the leading edge of online education is Udemy (UDMY, Financial), which has a crowd-sourced model that helps courses get to market faster.
Additionally, AI can be incorporated into classrooms and online education platforms alike to provide students with a more personalized learning experience, suggest relevant courses, provide feedback and more, significantly cutting back on the time that is wasted on reviewing familiar content and searching for gaps in a student’s education. For example, Coursera (COUR, Financial) recently unveiled generative AI features for its online platform, boosting its capabilities for grading, student feedback and lesson planning.
