Here are four lesser-known markets which AI will revolutionize within the next decade.

The applications of AI are continuously sweeping across all industries, with some adopting more technologies than others. Everyone talks about search, EVs, chatbots, recommendation systems, ads, and other spaces which have more visibility in tech. But what about other areas which people should probably talk more about?

For these hidden markets, I’ll evaluate the growth potential and impact of disruption to really underscore the excitement which should be brewed.

Game Publishing for Indies

A somewhat niche market is the publishing space for indie game developers. Look — we’re talking hidden potential here. Indie games comprise the majority of all existent video games found online and through platforms/consoles. 95% of all Steam games are indie-based, but only 40% of Steam sales are from this chunk. Current and aspiring indie developers struggle to get started and reach the finish line — part of the reason is the lack of any good publisher, willing to give them a boost in marketing, game design support, or funding. Additionally, 70% of commercial indie games are considered failures — they don’t generate any revenue. Only 30% of all games are considered financially successful — but of the 30%, only 7% of those games will generate enough revenue to fund a second project. Many indie games lack a good publisher, and many need the right publisher who can fit their needs. Many other indie studios also struggle to understand what they need out of the publisher — match-making is a necessity.

This is where AI comes in — by taking in the input of gameplay, music, writing, content, and other additions from the studio, a perfect marketing plan or match to a indie publisher could be calculated. We have enough data of past failures and struggles among game developers, and its up to deep learning to understand what’s best based on a studio’s circumstances.

Music Generation

AI-generated music is already a thing. The next step is for Music producers to begin mass adopting the power of AI to aid their creations.

Music production at its core involves in-house instrumentals and vocals, and/or digital development. The methods which exist today have never made it more easier for composers and producers alike to learn and create.

Music as a whole hasn’t been prone to AI-related projects. Just look at Google Magenta’s MelodyRNN and Adobe Creative Cloud’s VoCo — the latter of which has remained unreleased. The former, built as a “recurrent neural network” — is perfect for training and outputing sequential data like music. At the lowest level — the next note for a musical piece could be literally predicted based on the flow of previous notes… and entire melodies could be outputted like any other inferencing model. So why haven’t these two projects been successful?

They were both released around 2016 — a time where AI hasn’t reached its applicableness as it has now. The models were still fresh in theory, and society doubted its creativity.

Well, thanks to the likes of ChatGPT and other LLMs, do we still doubt the creativity of AI now?

Obviously AI-generated music still has a long way to go, and would never top the imagination of humans. But the use of AI tools to aid in production has never made it easier than today. Use cases include building solid foundational frameworks, playing around with others’ works to create parodies or remixes, and to learn potential next notes or beats for a given staff. That’s not even mentioning the current use of AI in generating music video content using UGC, alongside other areas.

And if you still aren’t convinced, just check out these 9 AI music generating start-ups/platforms. Who knows — some of your favorite artists could already be leveraging the latest AI technologies.

Public/Civil Infrastructure

The whole realm of civil engineering holds potential for disruption, lest we forget AI is already sweeping the space.

There are thousands of different tasks which not only could be automated one day, but elevated thanks to AI intervention.

These include the design and construction of roads, buildings, bridges, and other infrastructure projects. This article by Isakova highlights the idea of “smart construction design” and “construction process orchestration” which can leverage AI-enabled software to estimate building costs and reports, schedule project timelines and resources, and even output the most efficient plans. What really sparkled a bulb was when I read her last part: “fuzzy systems” — a way of mimicking human thinking when it comes to construction design. Leveraging AI to calculate quality assessments of project ideas given a range of costs could be groundbreaking. Indeed, even the most niche edge-cases could be detected — we’re talking dangers of weather, fire hazards, dangers of the area, and other risks.

This article by Tarutal Ghosh Mondal and Genda Chen also highlights the power of CNN to automate and improve building inspection tasks. It also mentions the potential of structural health monitoring (SHM) reaching a ~$4 billion market by 2028. SHM deals with the necessary maintenance and condition assessment of civil infrastructure. If we were to ride this wave of growth while leveraging the heaps of data available to perform autonomous processing, field monitoring could grow even more accurate and applicable.

Video Editing

What if video editors — YouTubers, film studios, and other companies — could become faster at editing and outputting magic?

Some of the most popular video editing software in the world are already adapting new techniques to enhance the editing process. DaVinci Resolve, Premiere Pro, and Power Director are all adding features that make an editor’s life easier: automatic cuts, quality/framerate re-adjustments, and other tools. But what about AI?

Well, projects like Synthesia and Lumen5 are already revolutionizing the space. AI avatars and voices are some of the interesting features from the former, which could easily make videos more personable. Lumen5, on the other hand, is more focused on AI-generated presentations which leverages ML models. These are used to predict the best final template and finishing touches to your product based on the video content as input.

Other AI effects across the other projects include automatic noise reduction and green screens, text-to-video conversion (inputting text to create fun animated videos), and video script analysis.

Go ahead and check them all out here!

That’s it for now — thanks for reading. But follow me as I post regularly on a weekly basis! Connect with me on LinkedIn and Twitter as well for more of my writing, product, productivity, and technology-related posts!