The Business Value of Artificial Intelligence - A guide for transformation

Aleksandra Hadzic
7 min readNov 16, 2021

“The secret of getting ahead is getting started”
Mark Twain.

In today’s dynamic and highly competitive business environments, getting started means accelerating the adoption of cognitive computing technologies quickly and in a way that delivers business value.

Today’s leading organisations realise that cognitive computing solutions are the key to future business success. This is evidenced in the millions of dollars being spent by technology investors on AI startups.

The most successful companies today, in various industries, are leveraging AI in their organisations, including:

  • The adoption of AI in business processes
  • The deployment of AI applications
  • The implementation of cognitive solutions

Cognitive computing solutions not only automate processes but also enhance employee performance and streamline operations. Cognitive computing solutions also enable business organisations to analyse and visualise data, predict outcomes, and scale operations.

However, organisations face three significant challenges when integrating cognitive computing solutions into their business operations:

  • The knowledge gap — Employees need to acquire new research and technical skills in AI
  • The skills gap — Employees need to obtain new skills and capacities in AI
  • The adoption gap — Employees need to adopt new ways of thinking and solving problems

Moreover, businesses need to scale their AI initiatives while maintaining high quality rapidly.

Who is shaping the future of artificial intelligence?

The goal of AI is to understand things.

This is different from the goals of machine learning and computer vision. The purpose of AI is to generalise, to learn what things, in general, look like, not just how to process examples.

When an image is distorted, AI algorithms either fail to understand what’s happening or ignore it. But the goal of AI is to understand, and understanding means understanding not just one distorted image but many distorted images. AI algorithms that understand, therefore, can reason about distortions. Rigid algorithms, on the other hand, repeat the same behaviour every time.

Part of the problem is that AI algorithms are fragile.

They are good at processing examples and using those examples to predict what will happen in the future. But they can’t reason about abstractions. If you tell them to draw a line, they will mark a straight line, but if you ask them to draw a line that varies a bit, they will draw a straight line, but they won’t know why.

One of the reasons is that AI algorithms are — just algorithms, and they don’t understand what they’re looking at. At Google Brain, AI pioneer Andrew Ng understood this, and they tried to solve this problem by training their algorithms to understand things.

The trouble is that training is very hard. When you train a neural network, you give it a set of examples and tell it how to calculate the probability its answer is correct. That almost works if the examples are chosen carefully. But in practice, the examples are chosen less carefully. So the AI algorithm ends up learning probabilities instead of understanding.

The second problem is the computational cost of training. With neural networks, the computational cost grows exponentially with the number of training examples, making it very expensive to train AI algorithms on large datasets.

Following this, it’s easy to understand that the business value of AI won’t come from AI itself — at least not at first.

The businesses that first reap the most significant rewards from AI are the ones that can harness it by making better use of existing data. AI will let us do things we couldn’t do before, but it won’t let us ignore our already existing data.

For example, after being a co-founder of Google Brain in 2011, Andrew Ng also co-founded deeplearning.ai and Coursera. The approach was to make AI easier to use. For instance, deeplearning.ai provides AI tools that are simple enough that anybody could use them.

Andrew Ng has also put AI to work, solving the problem of making it accessible to more people. The algorithms that deeplearning.ai uses in teaching AI are freely available, and anyone with an Internet connection can download them and use them for training their own AI programs. Ng has estimated that deeplearning.ai’s training algorithm on convolutional neural networks has been downloaded 50 million times.

Manufacturing AI gap

However, it seems like there’s more work that needs to be done. Google Brain’s former founder has launched a new company, Landing AI. On LinkedIn, it appears that the company has more than 70 employees. Investors in Andrew Ng’s startup Landing AI have given him $57 million, making him one of the most prominent figures in artificial intelligence.

In today’s competitive manufacturing environment, enterprises look to adopt new technologies quickly to stay relevant and profitable. Industrial companies also face the same challenges as more traditional companies when it comes to adopting new technologies. Training, change management, and data preparation are just a few of companies’ hurdles when implementing new technologies.

With the help of Landing AI’s enablement and transformation programs, customers can reap the benefits of AI. The company’s core product is LandingLens — an AI platform for industrial customers to build, deploy and scale AI-powered visual inspection solutions.

According to AI Magazine, billions of dollars have been invested in AI startups over the past decade. Artificial Intelligence (AI) startups are predicted to receive $90 billion in funding in 2021, up from $60 billion in 2020.

The only thing driving AI adoption is the need to deliver a better customer experience. AI can help deliver better digital experiences. Whether it’s AI as an enabler or an AI developer, AI enables many digital experiences. Whatsmore, AI-enabled technology can improve the customer experience by 80%. The customer is willing to pay for that.

One good example of consumer willingness to pay for AI is a virtual assistant called Alexa from Amazon. It’s being used less by consumers than one might think. And it’s not yet clear whether Alexa can become as ubiquitous as, say, Google Now. But it is believed that Alexa is “some of the leading stuff” that Amazon’s done.

Even among consumers, however, AI’s value proposition is not obvious.

According to App Annie, since its launch in 2016, Apple’s Siri has been downloaded more than 300 million times. But usage statistics are hard to come by, and it’s difficult to tell if those are people who use Siri rather than those who downloaded the app.

AI’s value proposition is best understood in the context of how humans behave. Therefore, we should go inside and trust our intuition when shaping AI and deciding if it’s a helpful tool. Is it all about AI optimising for efficiency and trying to mimic human behaviour? Let’s see what our intuition says about that.

The value of AI is being realised but in small amounts. But, as more AI becomes part of people’s daily lives, we expect to see sizable financial benefits. According to a McKinsey study, AI is expected to create $13 trillion of realised value to the world’s economy by 2030. For consumers, AI improved ad personalisation and recommendations, for example, and scene recognition in cars. For businesses, AI is improving inventory management, fraud detection, and customer service.

We are witnessing a historical evolution of technology, data, and talent.

By harnessing the business value of AI, we can power the transformational changes and deliver value to our customers and partners. This funding round allows us to accelerate our mission to bring AI capabilities to the masses and accelerate our leadership in the fast-growing AI solutions market. The emergence of AI is forcing businesses to reimagine their operating models — and, as a result, is fueling a wave of innovation and disruption across many industries.

How data-centric is manufacturing becoming?

The ability to leverage data, analytics, cloud, and AI technologies is critical to surviving and thriving in a digital business landscape. However, the methods in which these technologies are being deployed can vary widely, resulting in roadblocks to achieving corporate goals.

As a leader, how can you overcome this?

First, you have to understand that modern manufacturing and business are digital businesses. Your company depends on customers, suppliers, and partners for goods and services no longer relevant. How many digital connections your company has and how quickly it can respond to changing conditions matters.

Second, you have to realise that your company’s digital readiness is a journey. While you won’t reach 100% digital readiness overnight, you can undoubtedly achieve meaningful progress.

Finally, you have to realise that it’s not about technology, and it’s more about people, processes, and company culture. By focusing on the right areas, you can achieve transformation.

Got some questions about this topic? Feel free to contact me!

In the meantime, can also connect with me on LinkedIn.

‘Till next reading ,)

--

--

Aleksandra Hadzic

Researching AI. Merging Data Science and Digital Marketing.