The power and flexibility of AI mean that any industry can benefit from it in some way. Larger research-intensive sectors, such as pharmaceuticals, are among the most rapid adopters because of the huge cost-savings that AI enables. However, even the smallest company can use generative AI software, such as ChatGPT or Bing Chat, to answer questions to do with its activities. Here then is a look at some of the major applications of AI in business.
Retail
Amazon Fresh stores began opening in the UK – mainly in London – in 2019. They enable customers to choose what they want and exit the shop without stopping at a till. Cameras track what is picked up off shelves and sensors around the store work out who is leaving with what and charge it to the relevant Amazon customer account. However, Amazon has curtailed its ambitious expansion plans for 260 UK stores and in fact closed three of them in London in July. So, interesting as this technology is, for the moment AI is having more of an impact behind the scenes in the retail sector.
One example of this is demand forecasting, which uses machine learning to optimise inventory levels, and reduce the need to sell excess stock at reduced prices. And retail businesses, such as supermarkets, are using AI to get real-time data about their margins, enabling them to adjust pricing more aggressively to stay competitive with online rivals.
Larger retailers are using 3D imaging and spatial analysis to gain insights deriving from how customers behave in their shops that enable them to amend store layouts to present goods more enticingly. And AI is helping them to decide whether a given product will sell better in store or online. Hence the frustrating experience of popping into a shop to buy something you saw on its website only to find it is not there.
Sales and service
It is becoming difficult to find a customer service website that does not offer a dialogue with a chatbots. Like a new member of staff who might not be a great deal of use to start with but grows into a role to become a valued employee, chatbots are becoming more responsive, and can normally handle basic enquiries effectively. Chatbots use artificial intelligence, machine learning and natural language processing to work out what is being asked of them and reply accordingly.
Chatbots can also be used to increase sales by suggesting products to customers based on what they have shown interest in. And they can gather information from customers about their needs with a view to marketing products to them in future.
They can be equally effective in delivering services, assisting customers with booking flights or hotel rooms, and making other travel and entertainment plans. And their translation abilities are useful for companies that trade internationally and spare the need to recruit bilingual staff.
However, even though they are always working and don’t need to sleep or take a break it is not a good idea to get rid of your staff and let the machines rule. It is better to see them as a way of supporting employees by freeing them up to do more complex tasks and to focus on more valued customers.
Legal, accounting and professions
Software that uses natural language processing (NLP) is able to transcribe dictated letters and documents more quickly and cheaply than someone typing them can. The systems learn to distinguish between individuals and get to know their speaking style, further improving accuracy.
A more advanced application is predictive coding, which ranks the relevance of documents fed into it. This is particularly useful for lawyers who have to sift through large volumes of paperwork without knowing where nuggets useful to their case are buried. AI is also used in practice management to log hours spent on client work and take care of billing and manage workflow. Even contract production can to some degree be automated and, even better, the software can detect clauses that are needed but have not been included.
Other tools improve searches on law databases used to build cases and analyse previous ones to assist lawyers in creating case strategies.
The ubiquitous chatbots can help the public by acting as free online lawyers answering questions about routine legal matters.
Manufacturing
While robots have been toiling away on the factory floor for many years, their capabilities have begun to increase rapidly. AI-powered robots can perform more complex tasks, as well as packing orders and lifting heavy parts, and their ability to learn makes them adaptable. They can also self-optimise to make production processes as efficient as possible.
AI is well suited to quality control and will usually outperform humans in spotting flaws and faults. This ties in with predictive maintenance, where sensors monitor equipment and processes and send the data to AI software, which uses it to spot anomalies and underperforming parts that may be about to fail.
Similarly, digital twins, which are virtual copies of real-world systems, enable repairs to be made before equipment fails. This is particularly useful for remote applications, for example, railway switching gear, power grids and in the oil and gas industry.
Online security
Machine learning can block potential online attacks by comparing normal user activity with that of hackers. For example, excessive log-in attempts will trigger software that blocks that user’s access. Other AI technology takes an overview of a company’s network, reviewing all activity for unusual or suspicious activity in a way that human IT staff simply would not be able to, especially by detecting patterns that require large amounts of data to be processed before becoming apparent. Furthermore, AI can intervene autonomously to block potential threats.
Endpoints, that is the likes of laptops and smartphones, are vulnerable parts of a company network and can be protected by AI software that analyses how they are functioning to detect possible threats. It can block the endpoint from unauthorised access and protect sensitive data.
Another way AI keeps networks secure is by scanning them for vulnerabilities and applying patches to those vulnerabilities. It can also recomnend fixes and security updates when needed.
In addition, AI is sweeping through the following sectors.
- Social media. Brands are finding AI very useful for automating routine marketing tasks, such as posting and generating content and captions. In addition, it can be used to segment audiences and optimise advertising. It also important in content moderation, which would be prohibitively expensive if done by humans.
- Medical diagnostics. AI is helping clinicians achieve better outcomes for patients by making screening more accurate, thus spotting diseases earlier. It can also enable more personalised dosages and treatments, for example, of radiotherapy. AI-powered robots can help with surgery. And they are becoming crucial in analysing scans by detecting subtler patterns than the human eye is capable of.
- Research and development. Generative AI can identify fruitful pathways for research with minimal financial outlay, based on historical data, published scientific research and market trends. Drug development, which often requires vast amounts of research, is a key area for this, as are clinical trials. AI can conduct experiments, as well as assist in their design.
Of all the areas covered it is likely that R&D will change the most because of AI. As well as AI’s huge data analysis capabilities it is valuable for its capacity to integrate data and thus derive insights from it, in ways that have not been easy for larger companies with numerous sites and areas of activity. It is making research more efficient and cost effective, speeding up the rate of scientific discovery and technological advance, with benefits for all.
