What’s in store for business with the AI revolution?
To some extent, you can answer this with knowledge of the processes that companies are using to build the power of LLM’s and related technology into their workflows. There’s a lot going on, but there are also defined trends that help us to make sense of it.
Let’s go over some of the top findings that people come up with when they look at practical applications of AI in enterprise.
Revolutionizing Search
One of the fundamental building blocks here, according to a lot of experts I’ve heard from, is the application of AI to data searches
Some people talk about how this could be applied in Internet search engines. But others look to enterprise use cases like taking bank information – loan documents, mortgages, contracts, etc. and running the data through AI to get results that humans never could’ve come up with on their own. For example, if AI analyzes tens of thousands of bank documents and sees that people named Jerry tend to have higher balances, that’s one of those obscure pickups that human teams simply could not have found.
In a way, you could say we’re moving (hence the title) from quality assurance or “QA” to “Q&A” where machines give us answers to our own questions.
Data Pipelines
Other people with real knowledge of these industries are talking about the extraction, chunking and embedding of unstructured data into new data processes.
Usually, you’re coming up with a sort of flowchart design on a dashboard, to show how this is going to be applied all the way through a business, or at least across multiple business processes.
No matter what it looks like, this is a big one in entrepreneurial application of AI.
Leveraging Multiple AIs
This is another big one where companies are taking two different flavors of AI, and combining them for more effectiveness.
On the one hand, you have classic AI/ML which is more geared towards leveraging big data in statistical ways.
But then you have deep learning – those more intense applications where AI is starting to become more and more like a human in its cognition.
Can you use those more service-level things with the deep learning resources? Absolutely. Some workflows may respond better to one or the other of these models. So leveraging both of them really make sense for a business.
Human Augmentation for High Level Tasks
Here, we’re kind of talking about the human in the loop, and AI assistive process. It’s the idea that humans are getting help from AI to enhance what they are able to do in a particular workflow.
In a recent panel at IIA, Dave Blundin talked about how the people at Tailbox came to him and showed him his own voice giving a tour, in a way that, he said, was more capable than he himself could have done!
Talking about the convergence of technologies, Blundin made the point that it’s a wide open field now, and there’s a lot that we can make today that we couldn’t make before. Some of that, a lot of it in fact, involves that aspect of human augmentation.
Digitizing Conceptual Spaces
Some companies are also operating on a broader level to digitize what used to be manual operations.
For example, take a look at the company Retrocausal that uses this sort of approach toward assembly lines. “Digitize your entire line,” the company suggests, with big clients like Honda, Nvidia and Siemens (and don’t forget a U.S. federal agency: NASA!)
Data Flywheels
We also got some interesting input from another participant in the panel when it comes to using specific tools to guide enterprise.
Take a look at “the Engine” – pioneered by MIT right here at home, this concept furnishes startups with help on things like multimodal input and generative design. This is yet another big push related to rolling out some of the best new applications for AI technology in general!
So those are some of the takeways from recent panels on AI in business. Think about these types of use cases and how they may relate to your company!
