In Part 2 of our AI and Business series, Kinetic IT developers Domenic Horner and Daryl Crosbie talk about what you need to consider before implementing AI in your business (Read Part 1).
It’s been less than two months since we released our insight analysing the top ways artificial intelligence (AI) is transforming business. However, the AI race continues to progress at lightning speeds. OpenAI released ChatGPT-4, its most advanced system yet, with the ability to generate meal ideas based on what’s in your fridge, create a detailed learning syllabus, and even diagnose medical conditions in animals and humans.
Meanwhile, Microsoft introduced its new 365 CoPilot, an AI assistant for Microsoft 365 applications, and Security Pilot, an AI-powered security analysis tool that processes and responds to threats in minutes. In March, Bill Gates said “AI is as revolutionary as mobile phones and the Internet” and it “will help empower people at work, save lives, and improve education”.
However, in response to the full-throttle AI advancement, the Future of Life Institute has started a petition calling on all AI labs to “immediately pause for at least 6 months the training of AI systems more powerful than GPT-4”. At the time of writing, the petition has garnered over 22,000 signatures in three weeks, including global industry leaders such as Apple co-founder Steve Wozniak, Tesla CEO Elon Musk, Stability AI CEO Emad Mostaque, and Geometric Intelligence founder Gary Marcus.
The petition is not calling for an overall pause on AI development, but recommends AI labs use the pause to develop shared safety protocols for AI development that are audited by independent experts. OpenAI also released a statement regarding artificial general intelligence stating that “it may be important to get an independent review before starting to train future systems”. OpenAI’s overarching mission is to ensure that AI benefits all humanity and we need to be confident these systems will have positive impacts and manageable risks.
RELATED CONTENT: 6 powerful ways AI is transforming business
Should you implement AI in your business?
If you’re thinking about implementing AI in your business, you’re not alone. As AI continues to make waves, more businesses are seeing it as a vital tool to remain competitive. A report found that 83% of businesses say AI is a strategic priority and 84% say that AI will help them obtain and/or maintain a competitive advantage. A whopping 75% of organisations believe AI technology will help them move into new ventures.
While AI development races along, it’s pertinent for everyone to take their own pause before diving into AI integration. It’s clear that AI is the future – but implementing AI in your business is easier said than done. An Economist Intelligence Unit study found that 79% of executives believe AI will make their jobs easier and more efficient, while 37% of executives say the main obstacle to adopting AI is a lack of understanding of how emerging technologies work. AI technology promises enhanced productivity and output, but integrating these systems is a daunting task that requires plenty of forward planning.
We talked to Kinetic IT developers Domenic Horner and Daryl Crosbie about the top things to consider before implementing AI in your business, so you can get the most out of it.
1. Can AI help your business achieve its goals?
The most important consideration before implementing AI in your business is whether it aligns with your business strategy and will contribute to achieving your long-term goals.
Senior Applications Developer Domenic Horner says “Organisations will need to identify the areas where AI can provide the most significant benefits, such as automation, enhanced customer experiences or data-driven decision making.”
2. Should you build your own AI software or use existing tools?
Businesses need to consider whether the greatest value comes from bespoke AI software or existing tools.
Solutions Developer Daryl Crosbie says “AI software can be difficult to build as there are many aspects to consider, test and evaluate. You need to consider the model being used to train the bot, evaluate the training data, and adjust the model. The most difficult part tends to be acquiring enough diverse data to train the bot. With all this in mind, if there is existing machine learning software that suits your needs, it may be best to test and evaluate this first.”
RELATED CONTENT: 3 ServiceNow AI tools revolutionising IT service delivery
3. Do you have the infrastructure and talent to implement AI in your business?
If you’ve ever adopted a new system, tool, or strategy in your business, you already know it can be a lengthy and complex process. The same goes for integrating AI technology. You’ll first need to assess your existing IT infrastructure, including computing power, storage capacity and network security, to determine if it can accommodate AI systems. If there are gaps, you may need to consider upgrading your infrastructure or migrating to cloud-based solutions.
While it’s crucial to identify appropriate tools and software to manage and monitor AI solutions, you’ll also need the right people. AI implementation requires skilled teams of engineers and experts with AI and machine learning skills and knowledge.
“You may need to hire new talent, upskill your current workforce, or partner with external providers,” says Domenic. “You’ll also need to invest in development and training to stay on top of this rapidly changing technology.”
4. Is the AI technology compatible with your current processes?
Your current IT infrastructure and teams may be well-equipped to implement AI in your business – but will they be compatible with AI technology? You’ll need to plan how AI systems will integrate with your existing infrastructure, software, and processes.
“Work with your team to identify potential compatibility issues, such as data formats and communication protocols, and address them early on to avoid headaches down the road,” says Daryl. “By designing APIs, adopting standard interfaces, or leveraging middleware solutions, you can achieve a smooth and seamless AI integration.”
5. Is the AI technology scalable?
If you’re doing the work to adopt AI in your business, you also want it to be scalable and adaptable to changing needs. As your business grows, you’ll need your AI systems to scale easily to handle increasing data volumes and users and expand across business units or regions.
“Flexible, modular, cloud-based AI solutions are best for scalability,” says Domenic.
RELATED CONTENT: Task automation: 9 steps to mitigate the security risks
6. What data was used to train the AI system?
Not all AI tools are created equally. AI systems rely heavily on data for training and functioning, so it’s crucial to have high-quality, diverse data to train AI models effectively. This is one of the greatest challenges in implementing AI in your business, as obtaining and cleaning data can be time-consuming and expensive. You’ll need to examine the quality, quantity and variety of the data, and check for data accuracy, consistency, and relevance, to ensure it’s suitable for AI implementation.
“The data used to train the AI model is very important, as it needs to be as diverse as possible,” says Daryl. “If it contains biases, the algorithm will likely hold this bias in its results. There are numerous models that can be used to build AI applications. Each model can have its pros and cons when it comes to the data being used. Some machine learning models can produce data that can be difficult to interpret, so depending on the type of data you wish to work with, and your needs for the data, choosing the right model should be given consideration as it can be very important.”
7. How effective is AI technology in your business?
AI tools are powerful, efficient, and even entertaining – but they’re not always the right fit. AI tools aren’t infallible, and they can make mistakes – just like humans.
Domenic says: “AI-powered tools like the language model ChatGPT can help with things like writing articles or work instructions, providing highly technical answers to questions, and triaging IT incidents. However, there are downsides to this as the accuracy of the answers are not validated by any human and they can, and sometimes do, contain incorrect information.”
One way to overcome this is by starting with small pilot projects to test the effectiveness of AI solutions and their integration into your systems. Your teams can monitor the results, gather feedback, and identify areas for improvement. This will help you refine the AI models, minimise risks, and maximise the benefits of your AI tools and systems.
8. How high are your expectations when implementing AI in your business?
An iterative integration process will also help manage business expectations around AI. While AI has made extraordinary advancements in the last few years (and months!), many AI tools are still finding their feet in terms of capabilities.
“The technology in AI and machine learning is advancing at a very rapid pace and there’s more and more use-cases being developed and proven with each advancement”, says Domenic. “However, models like ChatGPT are still very fresh and there have not been many implementations of this technology to learn from.”
If you’re ready to implement AI tools in your business, be sure to set realistic expectations. AI is not a magical cure-all and can’t do everything and fix every problem – yet. AI has already shown itself to be a powerful tool for solving complex problems and improving efficiency, and as the technology continues to advance, the possibilities for positive impact will become almost limitless.
RELATED CONTENT: Case study: Qantas welcomes aboard AVA
9. What are the financial costs of implementing AI in your business?
Like any new technology, AI tools and systems can be expensive.
Daryl says: “Due to the cost of integrating and running AI to perform tasks, businesses will need to evaluate the return on investment carefully. However, most implementations of AI stand to significantly reduce costs to the business for the long-term.”
Domenic agrees: “AI-powered technologies can automate repetitive, time-consuming tasks, freeing up employees to focus on higher-level, more creative tasks, leading to improved efficiency and productivity for businesses. For example, instead of using human talent to answer simple customer queries, you can hand the job over to an AI-powered customer chatbot. They’ll get it done faster and free up time for people to manage more complex tasks.
When you’re considering the total cost of implementing AI in your business, consider the costs of hardware, software, training, talent acquisition and ongoing maintenance. You’ll also need to compare the costs with the anticipated return on investment (ROI) to ensure the project is financially viable. Many businesses discover that AI tools can vastly reduce time and resources, making it a more economical option.
10. What are the ethical implications of using AI?
While AI has the power to revolutionise every industry, it’s important to acknowledge and mitigate the risks and ethical concerns that come with the development of this technology and implementing AI in your business.
“There are some ethical concerns as the AI must be trained using publicly available information which has the potential to contain Licensed or Protected information,” says Domenic. “There are also concerns about privacy, security, accountability, transparency, and potential biases. As AI becomes more widely adopted, we can expect to see increased demand for regulation and standards to ensure that AI is developed and used in a responsible and ethical manner.”
With greater awareness and education, organisations will be better placed to navigate the challenges and opportunities of AI.
11. Is the broader business prepared to adopt AI?
Implementing AI in your business can result in significant organisational and cultural shifts. A change management strategy can help employees adapt to new tools, processes, and ways of working. You’ll need to develop communication, training and support programs and be prepared to address concerns and manage expectations throughout the AI integration.
And if you need assistance developing all these strategies, programs, and communications? AI can help with that – just ask ChatGPT.