How machine learning is boosting businesses

Artificial Intelligence: a quick definition

In general, artificial intelligence, or AI, refers to machines performing tasks that previously would have required human thinking; for example, pattern recognition, data synthesis and analysis, and forecasting from current (i.e. limited) information.

Factors driving the coming AI breakthrough

“The ingredients for a breakthrough are in place. Computer power is growing significantly, algorithms are becoming more sophisticated, and, perhaps most important of all, the world is generating vast quantities of the fuel that powers AI—data. Billions of gigabytes of it every day.”

These and other drivers are behind the growing use and influence of AI in our digital development. Another factor is finance. Even if a lot of investment is a case of jumping on the AI bandwagon, there are still impressive quantities of money flowing into AI research and development. MGI cite a 2016 figure of total investment in AI of $12 billion, with 60% of that amount going to machine learning.

However, despite the well-known cases of Google and Amazon, we’re yet to see widespread commercial adoption of artificial intelligence. The MGI research drew on 160 global use cases from a range of industries but only one in eight projects had left the experiment stage behind. Outside of Silicon Valley (or more accurately, outside of the tech sector) there are few companies embracing AI for process automation and data-handling despite the technology being well-proven and subject to continuous improvement. We’re at the tipping point for AI but we haven’t quite ‘gone over the edge’ yet.

You might be also interested in the article:

Digital Product

Predictive maintenance for wind turbines - an interview with Boldare’s machine learning engineers

That said, MGI have identified a six-point profile of companies most likely to invest in AI (i.e. the current early adopters):

• Forward-looking – companies already investing in related technology are more likely to invest in AI and machine learning.
• Resources available – investors tend to be larger companies.
• Broad scope – early adopters tend not to specialize in just one type of technology. They use multiple AI tools for multiple needs.
• Focused – the AI investment is linked to the company’s core business, it is more than a case of exploring an interesting side road on the company journey.
• Creative – such companies are using AI for product and service innovation, not just to automate repetitive, rules-based processes.
• Committed– there is a strong pro-AI leadership at the c-suite level.

Outside of companies like this, investment remains cautious (MGI found only 20% of companies are planning to increase AI spending by more than a tenth) but these early adopters are leading the current wave of interest in artificial intelligence.

Potential benefits of investing in AI

But what exactly is driving these early adopters. What benefits are on offer with AI and machine learning?

Accuracy

Technology is approaching a point where it can take over certain tasks from the human brain. However, unlike the human brain, AI doesn’t get tired and this has an obvious benefit for accuracy and error rates.

Profits & financial performance

Some of the big names leading the AI vanguard in the commercial sector have shown its financial viability. Amazon’s purchase of Kiva (a robot ‘picking and packing’ company) led to more than a 75% reduction in packing times and a 50% increase in inventory capacity; meanwhile operating costs fell by 20%.

Forecasting

Better supply chain forecasts result in the design of a better offer, specifically tailored to the market. The MGI report suggests that in the retail sector, sales lost due to product unavailability can be reduced by up to 65% thanks to AI-powered forecasting.

Production

Production processes can be optimized using AI; for example, ‘collaborative robots’ (i.e. capable of interacting intelligently rather than simply carrying a rigid sequence of movements) can potentially increase productivity by 20%.

Marketing

Promotion of goods and services depends on offering the right package to the right person with the right message and the right price, at the right time, and that depends on being able to use the available data to its fullest advantage.

Delivery

A great customer experience should offer value, be personalized, and lead to increased revenue. AI-driven ‘monitoring’ of customer activity can be used to, for example, make real-time purchase recommendations. And new smart payment systems mean the customer’s purse or wallet doesn’t leave their pocket (though the money WILL leave their account!)

AI and Machine Learning at Boldare – a real application

At Boldare – the digital product design & development company - we see artificial intelligence, and specifically machine learning, as critical to the future of product development and often central to the digital transformation journey of our clients. We aim to create machine learning models that can automate tasks, improve decision-making, optimize costs and processes, help to understand customers, and boost profitability.

One example of boosting a client’s business with machine learning tackled the problem of content moderation. Whether it’s profile pictures, forum posts, or in this case, automobile tires, as content is added to a system, checking that content is a tedious, repetitive task. One of our projects was creating and maintaining an integrated system for a tire wholesaler. As new products are added to the system, each one must be checked and that checking process is time-consuming.

By collecting the product data and a set of manual moderation in an external database, we could train a machine learning model to predict the probability of the data of an unverified product being correct. The AI’s predictions concerning new products were then compared to the results of the manual process until we could guarantee an agreed level of accuracy, leading to fully automatic moderation of the system’s content.