It’s impossible to fully follow and understand the artificial intelligence tech landscape these days. It’s all we regular mortals can do to read a bit about it and let it roll over us like a tsunami. One latest development that’s possible for me to understand in a way is the recent abandonment by Open AI’s consumer app, Sora designed to allow users to place themselves in videos along with characters form Disney’s Marvel and Pixar studios. The company decided that the product not only wasn’t profitable, but every use required a huge draw down of the computer chips required to run it thereby depleting that critical resource for Open AI’s more promising and future profitable products. Furthermore, the Sora app having made a big splash upon release, was losing users fast. The idea that a product needs to be profitable and have a user base is an old-fashioned idea I can understand.

I’m a firm believer in the idea that technology develops along its own path and sometimes at its own speed, and that we may or may not see the end goal. That’s why any central or governmental regulation that preempts technology early on can be so inimical. Medical advances certainly show this kind of unexpected development where a new technique or drug explored for one purpose finds a use in another unplanned avenue.

Another related idea that I need to keep reminding myself of is that technology can start with a trivial application like making videos of yourself with animated characters, but end up being part of something useful that wasn’t the original plan. Video gaming—another area that just isn’t my cup of tea—eventually became a training resource for the U.S. Military. A widely sited instance of consumer product and military crossover is the Bradly Trainer of the 1980s, a modified arcade cabinet from the game Battlezone that was used to train Bradley Vehicle gunners.

Mechanical curiosities have a long history and were the stepping stones for the development of fundamental technologies. The word automata derives from the Greek and means “self-acting”, and describes mathematical models or abstract machines and how they process information. They are models of computation and date back to the ancient world when inventors created the first mechanical toys employing steam, water and the science of pneumatics to move objects.

These objects were still curiosities during later periods when Renaissance inventors like Leonardo da Vinci created his mechanical lion which used gears to open up revealing lily flowers! Leonardo designed the lion to honor French King Louis XII—it was an amusement for a very small and privileged audience not a product idea with a clear future. In earlier centuries the early market for such experiments were the princes and courts of Europe and were, I suppose, part of currying the favor of those in power so to win the chance to keep experimenting and learning. Today it seems that while the inspiration for new artificial intelligence products often is still amusement (seeing yourself alongside Darth Vadar), the initial user base has been democratized.

So Open AI shifted its priorities away from a new consumer application not because it was frivolous, but because it wasn’t looking like it would be profitable. Hurrah; some old-fashioned business sanity in the digital universe. Perhaps, the consumer video app could have been side-lined sooner. Perhaps not. But it does seem to call for examination of how we think about artificial intelligence and the world its creating.

While I’m vociferously against government preemptively regulating AI (as vociferous as one can be sitting alone in a room), we do want to help the technology develop along lines that create real value. History shows us that we can’t predict what routes those areas will take; they may seem frivolous at first. But, surly the future of AI is not with creating more diversions for an already over stimulated, over consuming society. Part of the future surely must lie with quantum computing and using computers to solve ever more complex problems that will contribute to a better life for their creators. We don’t really need to part of the digital world dancing with CGI, but we do need to make the digital world part of ours to solve disease, hunger and the problems that have always been with us.

https://www.wsj.com/tech/ai/the-sudden-fall-of-openais-most-hyped-product-since-chatgpt-64c730c9?mod=lead_feature_below_a_pos1
https://copilot.microsoft.com/chats/9CYsgSAdVCehd8X5NNDgf