Will the Democratization of Technology Accelerate Progress in AI?

If you were to poll the computing industry today for “the most pervasive technology of our time,” I suppose artificial intelligence would easily top the list.

And for good reason – the last decade of advances in artificial intelligence has certainly been exciting. but the IMPACR From this innovation follows the principle of William Gibson: “The future is already here, it is not evenly distributed.”

The funny thing about AI in particular is that people believe that the success of AI should be evenly distributed. If Tesla can drive your car automatically and Google Photos can match your elderly parents’ faces to their children’s photos, why can’t your company increase revenue and reduce cost via AI? Heck, AI can’t even figure out how to load your spreadsheet stack into a data warehouse!

So, what is the reason for the disconnect between AI innovation and its impact? The issue is twofold. First – all computing challenges are not the same. While some exciting topics such as computer vision have made huge leaps in recent years, most traditionally painful business data processing problems still go far beyond the capabilities of modern AI. Second, the engineering tools and practices for the success of AI and machine learning are still in their infancy.

Today’s big tech stores are largely solving their data and AI problems by hiring armies of expert software engineers to “cluster” data pipelines together using bits of AI by hand. This is exacerbated by the disparate state of open source tools. Unless your company can hire a lot of Silicon Valley quality software developers, you’re out of luck. To democratize advances in AI, we need to do two basic things:

  • Focus on Human Interfaces and Artificial Intelligence: We need to acknowledge that in many places, AI cannot go to great lengths. Instead, we need innovators to focus on AI as a more From human labor, not a substitute.
  • Bring people together across skill sets: We need to understand that The democratization of technology needs to bring together groups with different skills. The next generation of AI tools needs to allow all target groups to do their work as they see fit, while sharing each other’s challenges and progress.

Today’s big tech stores are largely solving their data and AI problems by hiring armies of expert software engineers to manually “link” data pipelines together with pieces of AI. This is exacerbated by the state of open source tools. Unless your company can hire a lot of Silicon Valley quality software developers, you’re out of luck.

That is why, going forward, I see three main trends that will play an important role in the democratization of AI:

  1. data engineering: I expect that developer-focused interfaces like SQL and Python will become increasingly interoperable with low-code tools. As the software matures, cloud-hosted services will make this new technology very easy to adopt.
  2. artificial intelligence engineering: I expect MLOps to enter the Cambrian Explosion phase in 2022. We see them in the startup market as companies compete to solve the narrow parts of the overall AI engineering pipeline. Some of these startups will find highly valuable leverage points in these pipelines and gain traction quickly; Others will vanish.
  3. Low code and no code: I expect that the next generation of low-code, no-code applications will be able to function like “automated coders” using synthetic AI and software synthesis. Non-programmers will be able to create the moral equivalent of custom software without having to know how (or if) they do it.

Next year promises to be a very confusing time for AI, especially in areas like MLOps where the stack has not started to shake. Be sure to monitor the human AI interfaces that facilitate augmented intelligence with low-code and no-code tools. While tech news stories about the achievements of AI will continue to impress you with the possibilities, you have to understand that practical uses of AI in business will continue to be scarce.

.

Leave a Comment