Artificial intelligence (AI) transformed industries that have access to huge datasets. Highly customized algorithms analyze and interpret these large data sets. The most known examples of this success can be found in consumer-facing internet businesses like Google, Amazon, Netflix, or Facebook.
Over the last twenty years, companies such as these have grown into some of the largest corporations. While their growth is outstanding, it is overshadowed by the ever-growing volumes of data they consume due to our increasingly digitized society.
If AI is going to unlock the truly world-changing value that many believe it will, then businesses in other industries have to consider different approaches. Industries such as agriculture, manufacturing, healthcare, or logistics are not going to have millions (or billions) of people freely sharing volumes of their personal data, in the way that they do with Google or Facebook – the models of interaction between consumers and the businesses are completely different.
It’s generally been assumed that AI – and in particular, deep learning, which relies on complex deep neural networks – requires enormous volumes of data in order to surface the insights needed for genuine transformation. But what if we could get insights from smaller volumes of data?
This is the starting point taken by Dr. Andrew Ng, the founder of Google’s deep learning research group, Google Brain, and chief scientist at Baidu’s Artificial Intelligence Group, with his current venture, Landing AI. Also, the co-founder of the online learning portal Coursera, Ng is a former director of Stanford University’s AI Lab, and is widely considered a pioneer in the field of AI – in particular, deep learning.
As one analyzes industries like manufacturing and healthcare, the world of data is no longer built around homogenous infrastructure staples – web browsers, cloud servers, mobile apps, and standard operating systems. What does this mean for AI? It means that “plugging in” AI-as-a-service becomes less straightforward, and the customization that’s necessary becomes expensive.
The solution? Focus on the data, not so much the technology.
“I think machine learning has transformed the consumer software companies”, Ng says. “Google and Baidu have AI teams, also Microsoft, Facebook – but once you look into other industries, candidly, I’m not seeing the potential value of AI being realized yet today.” Yet shortly after deciding to take his mission to release the value of AI to legacy industries like manufacturing, healthcare, and agriculture, Ng realized that maybe he had been, in his own words, “a little naive.”
“That recipe that a lot of us had collectively built at [internet companies] – that recipe doesn’t work at all these other industries.”
“So at Landing AI, one of the things we’ve been doing is working out that recipe … I’m excited about that; it’s a big thing that those of us in AI have to figure out.”
Ng and the team at Landing AI are focused on the manufacturing sector. They recently completed a series A funding round, raising $57 million from investors, including IoT-focused fund McRock Capital and Insight Partners, Intel Capital, and Samsung Investment Fund.
They recently showcased their manufacturing-focused MLOps platform, LandingLens, an application that uses computer vision to identify defects throughout the manufacturing process.
In a recent interview, Ng mentions, “One of the things that excites me is taking tools that exist like supervised learning, and building the platforms that make it possible for there to be thousands or tens of thousands of unique neural networks for manufacturing.”
“I think this is an AI-wide problem – take healthcare for example – every hospital has a slightly different way of coding their records, so you can’t have a single monolithic neural network to process every single hospital record.”
So what does this mean for AI start-ups? This means customization and a lot of it! More importantly, AI companies need to effectively be consulting companies (and partners) rather than platform service providers capable of operating at scale.
In an article with Bernard Marr, Ng claims “My challenge is how can I help maybe 10,000 companies build and deploy machine learning models without having to grow Landing AI to have 10,000 machine learning engineers to do all this customization?
“The only solution is to build vertical platforms, which is what Landing AI is doing … platforms that are fast and easy … to enable the manufacturers’ [workforce] to be able to train and deploy their own AI systems, so we can then collectively tackle this very heterogeneous world of manufacturing and other industry sectors.”
As we approach an AI-driven world, the answer is to be data-centric. In order to do this appropriately, the focus has to be on engineering the right dataset, which feeds into an open-source neural net AI model to achieve top performance. In simple terms, the focus is to use high-quality data as opposed to high volumes of data. However, this requires data and domain experts to analyze such data so that AI can be applied to a healthcare facilities, manufacturing plants and more.
This concept is clearly different from the software-centric companies that have been the force behind the rollout of AI applications. As Bernard Marr quotes, “A step away from “big data”, perhaps, towards “good data” that accurately summarises the information needed to achieve insights.”
As I have worked with many companies dependent upon legacy systems (both B2B, B2C and B2B2C) that are ready to utilize their massive amounts of data, I see that a tremendous amount of effort is required to customize machine learning and AI tools.
Customization is required to leverage expert knowledge within organizations to focus on the data rather than the coding, modeling, and algorithms.