Upgrade Your Knowledge | USU Blog

AI Use in IT, Challenges, and Limitations

Written by Michael Schemel | May 27, 2021 7:27:12 AM

There are sensational narratives out there that describe artificial intelligence or AI as man's last invention. Other interpretations of this view are that AI is man's algorithmic savior. AI will increase human innovation, treat incurable diseases, and enhance human creativity.  

Such notions have given rise to AI solutionism, where this technology infiltrates every aspect of life and, to some extent, dominates society. AI, therefore, presents an existential danger to humanity.  

To quote Nick Bostrom, an AI specialist and Oxford academician, "Before the prospect of an intelligence explosion, we humans are like small children playing with a bomb. Such is the mismatch between the power of our plaything and the immaturity of our conduct."  

These extreme views on AI development cause hysteria, giving much fodder to AI-inspired blockbuster movies. Unfortunately, such inferences also handicap the real progress of AI development. Lofty ideas on artificial intelligence jeopardize its value to society, setting unrealistic expectations of its capability.  

When can AI be used in principle?

Artificial intelligence technology uses in business has grown by 270% in the last five years. In 2015, less than 10% of companies had any AI technology applications. In 2019, the number of organizations using AI rose to 37%, as per a Gartner report.

More businesses are boarding the artificial intelligence for IT operations or AIOps train. As a result, AI is disrupting the IT sector, taking up tons of workload in cybersecurity, help desks, and other crucial IT sectors.

Data by Tata Consulting Services shows that the IT industry uses AI technology the most, with at least 46% of businesses in IT incorporating AI in their everyday work processes. AI streamlines increasingly complex IT processes. It also helps accelerate the IT innovation process. In addition, AI technology can improve workflows in IT, increasing accuracy and efficiency.

So, what is the impact of mass use of AI in IT? Will IT jobs go the way of the railroad worker or switchboard operator as AIOps takes over? To answer this question, we should first describe what AI technology is and what it is currently capable of. 

What are the limitations of this technology?

Artificial intelligence simulates human intelligence, leveraging computer systems and machines. AI technology can automatically perform content learning, text and speech recognition, or problem-solving.

To perform such functions, AI tools need massive amounts of data and computing power. There are five main subdivisions of Artificial intelligence technology. They include; 

  • Natural language processing
  • Machine learning
  • Deep learning
  • Speech recognition
  • Image processing

The IT field naturally gravitates towards deep learning and machine learning. Machine learning parses data via special algorithms, generating the desired result as per the analyzed data. In Deep Learning, AI tools also use algorithms. However, deep learning tools process large amounts of data and classify it using characteristics such as text, sound, or images.

AI usage can be the strong artificial general intelligence kind that replicates the human cognitive abilities. It can also be weak intelligence that runs virtual assistant tools such as Siri. This is the most common AI in solving a single problem and executing its tasks to perfection. However, such AI barely has any human cognitive abilities and only works in a controlled setting.

Strong AI or Artificial General Intelligence (AGI) is currently a theoretical concept. An AI tool capable of language processing, computational functioning, and image processing via cognitive function is still a dream for the AI sector.

To create such systems, the industry will need a massive network of narrow AI systems working in partnership to imitate human reasoning. The world's most robust AI infrastructure, such as IBM’s Watson, takes at least 40 minutes to mimic one second of human neuronal activity.

SpiNNaker or Spiking Neural Network Architecture is the world’s largest AI brain. The over a decade old, University of Manchester AI tries to mimic the human brain neuron activity. The project initially ran its experiments using 500,000 core processors.

The EU's Human Brain Project AI now runs 1,200 linked circuit boards but is still incapable of simulating the functions of a human brain. According to Steve Furber, SpiNNaker team member and University of Manchester computer-engineering professor, “Even with a million processors, we can only approach 1 percent of the scale of the human brain, and that’s with a lot of simplifying assumptions.”

Currently, the application of AGI is in advanced robotics. SpiNNaker, for instance, runs the SpOmnibot, a robot that makes real-time navigation choices via vision sensors. Beyond Artificial General Intelligence lies Artificial Super Intelligence; this is in the science fiction domain. Such an AI would make rational decisions and even build emotional relationships.

IT challenges now and in the future

As it stands, the creation of an AGI requires resources far beyond most business capabilities. For this reason, AI support in IT does not threaten the future of IT professionals but is opening new career opportunities.

AI tools will completely automate data analysis, a sector that buries analysts under heaps of data as humanity increasingly creates more data. The AI advantage is its ability to process data at superhuman speeds, with more accuracy, and without fatigue.

AI will also automate the mundane tasks that humans were not good at in the first place. For example, humans are not good at large-scale pattern recognition or repetitive tasks; AI excels at pattern-based tasks. IT professionals, therefore, will need to transition and become AI trainers and monitors.

They will need to evolve and learn new skills in voice recognition technology or AI algorithm creation and management. Technical support will become crucial to businesses. As an illustration, by 2019, 10% of all IT hires were bot interaction scriptwriters.

In 2020, 20% of businesses had a fraction of their employees guiding and supporting their AI neural networks. also shows that AI and automation could create over 50% more opportunities in the IT sector in the next few years.

Source: [KTSDESIGN/SCIENCE PHOTO LIBRARY]/[Science Photo Library] via Getty Images

 

AIOps as a defensive stance

IT businesses also need to embrace AI use as a defensive stance. Larger companies that have adopted AI are looking to roll out their AI tools in various sectors. Google, Amazon, and Tesla will not stop rolling out AI products until they have transformed all business sectors.  

There are also rising hungry AI startups that will run the AI economy and, in the process, cannibalize much revenue from the leading AI bigwigs. Businesses that resist AI use will soon begin to feel the heat as their inefficiencies start to drain their profitability.  

Their AI using competitors will, on the other hand, reap the benefits of efficient operations. IT businesses that fail to adapt will need to cost cut and eventually die. So under-investing in your AIOps strategy is not operational efficiency but a business risk. 

How AI improves workflows in IT

In today's business environment, time is like gold dust, and there are plenty of deadlines. Your IT business, therefore, needs to fulfill its tasks faster for impressive results. In addition, the IT client brief has also become more sophisticated, making AIOps the perfect building block for a dynamic, diverse, and complex IT environment. Below are ways that AI can improve workflows in IT. 

Deep data insight generation and proactive monitoring

AI technology can research, monitor, and report specific data. AI tools are also perfect for predictive analytics, showing deep market movement and trends. They can monitor network traffic and enhance business decision-making.

Automation of low-level cognitive duties

AI can become part of your IT strategy by letting it manage your challenging but repetitive round-the-clock tasks. An AI desk agent, for instance, is the perfect emergency IT service assistant, providing 24-hour self-service options that you can scale up when necessary.

AI-enabled desk agents hold intelligent chats with clients and learn from these interactions, making them very adaptable. Consequently, your AI tools can enhance productivity and free your staff so that they can focus on innovation and crucial tasks.

Customer experience improvement

Your AI tools can support your personalized customer experience strategy. They can guide your clients to purchases and help them learn more about your products.

Automatic reminders

Your AI tools can also automate your project deadlines, meetings, customer service, and schedule management reminders. For example, you can use these tools to keep your customers up to date with payment schedules automatically. In addition, your AI tools can perform A/B tests and determine the most effective day, hour, or medium of communication for better outcomes.

Fraud detection

Combine the analysis of statistical data and AI tools to mitigate technology-related fraud. Your pattern-detecting machine learning tools can synthesize data, highlight fraud, and help your IT department set up proactive preventative measures.

How exactly can AI be used in IT in practice?

Development of secure systems

Cybersecurity has become a high-priority area for businesses. As a result, the need for cloud and security application development specialists is on the rise. Yet, as it stands, over 3.5 million cybersecurity jobs remain unfilled.

AI is now a large component of data and cybersecurity application development. Machine learning and advanced algorithms can create high security and better identify data breaches as well as threats.

Efficient coding

An AI tool can detect, test, and beat software bugs faster and more efficiently. They can help fine-tune your code for bug-free codes and applications. The efficiency of AI tools in coding increases productivity and cuts down code production time.

System testing and network maintenance automation

Rather than have a reactive IT network maintenance strategy, have a proactive approach that pairs AI deep learning with IT workers in backend automation processes. AI tools will save human hours spent on manual maintenance.

These tools will also learn and enhance their algorithms over time. AI tools can optimize a host server's responses and strengthen its operations.

Quality assurance

Your AI tools can provide vital support during the application deployment process. For example, they can automatically eliminate all software bugs and plug gaps in code.

What trends will come here?

Artificial intelligence for IT operations is now going mainstream due to its ability to streamline fast decision-making and IT efficiencies. Other top trends expected in IT use of AI include;

  • Increased adoption of artificial intelligence for IT operations as more businesses adopt third-party machine learning platforms such as Amazon Web Services.

  • There is a shortage of AI skills amongst employees, so the creation of in-house AI has taken a backbench. Businesses that do not have a robust cognitive industry or AI outlook could get overwhelmed with the process of building an AI from scratch.

  • IT organizations are facing increased complexity in operations. They, therefore, need increasingly aware and contextual AI tools that have enhanced self-learning algorithms.

  • Most of the current AIOps tools run from the cloud. However, as data volumes enlarge, running a cloud-based AI operation has become slow and expensive. Therefore, savvy IT businesses will use AI tools that run on the edge of their networks for cheaper and faster processes. In addition, such devices will provide real-time AI efficiency.

  • AIOps will provide more control and visibility of the IT environment to businesses. However, this aspect of AIOps will also cross the lines between personal and work time and data. For this reason, there will be increased privacy issues to consider. Therefore, it is important to have the legal and HR departments strike a balance between businesses' needs and user privacy.

  • Increased AIOps will shrink the data entry sector and improve application development and data science jobs. The data scientist will be vital in AI systems management and monitoring. As per Gartner data, 90% of digital businesses will require a data architect by 2025.

  • Governments are waking up to the fact that AI is the future and are investing in it. There will therefore be more research and development incentives and focus on AI.