Are You Prepared for the AI Revolution?
We are in the middle of the Artificial Intelligence (AI) revolution. It has become so prevalent, it’s ignited a veritable AI race between some of the largest tech vendors in the world. Microsoft has formed a key partnership with OpenAI off the back of ChatGPT’s success. Google reacted quickly with the announcement of several AI initiatives, such as Gemini (formerly named Bard), clearly feeling threatened as the model of AI could render search engines such as Google less important.
So, how does one make sense of AI for enterprise IT?
In recent months, we have seen blog posts and videos showing use cases for generative AI in many different roles and industries. HR departments leverage AI to generate job specs, software developers to generate code, marketing teams to create images and written content, PR and legal teams to generate press releases and legal documents, and more. There have even been cases of physicians providing AI tools a list of symptoms to get a patient diagnoses.
How Does AI Affect Enterprise IT?
There are some genuine AI features included in existing IT tool sets, such as Microsoft Copilot for Microsoft 365 and Copilot for GitHub. Several Independent Software Vendors have been quick to include AI in their marketing and feature lists. However, it’s questionable whether these contain AI at all. This is not surprising. Microsoft can move quickly on AI, as they have struck a key partnership with the flagship AI technology vendor: OpenAI. They have an advantage with the massive resources at their disposal to take the lead in the AI market. Other smaller vendors are under pressure to adapt AI as customers seek its inclusion in product sets, but they do not have the same resources as a Microsoft or Google. So, they have been developing piece meal features leveraging OpenAI APIs and other AI tools to augment their products to mixed results.
Generative AI is quickly becoming synonymous with AI. However, there are many facets to AI. What seems to be getting lost in the hype of generative AI is Machine Learning. Machine Learning can help automate repetitive IT tasks and learn from how employees in the organization use technology to optimize the experience and improve productivity. GenAI leverages large language models populated with data sets containing public information. Organizations can create their own local data set and leverage a GenAI service that can use that information without exposing the data online, which could be useful to have an AI service that can carry out tasks with information that is relevant to your organization.
One area where we will see rapid improvements thanks to Machine Learning and Artificial Intelligence is the automation of enterprise IT workflows. Automating the build of infrastructure will be mostly handled by AI. Automatically remediating issues and anomalies that occur in a day-to-day enterprise IT environment could potentially be handled using automation enhanced by Gen AI and Machine Learning capabilities. Forecasting growth and hardware requirements can be handled using Machine Learning, improving application performance and launch times could also be handled by Machine Learning. Of course, automating the management of your applications can be assisted with Machine Learning and Artificial Intelligence too.
If you hear this as an Administrator and feel a sense of dread this type of automation rendering your job obsolete, don’t worry.
I remember over 20 years ago, when cloud computing was becoming a buzz term, industry analysts and mainstream media was predicting a decimation of tech jobs. The reasoning was, when infrastructure was moved to the cloud, there would be less overhead to manage and maintain the infrastructure going forward. Of course, fast forward to present day, and it turned out that the successful implementation of infrastructure in the cloud has led to the creation of jobs and opportunities for existing IT professionals to upskill for cloud-related work. Unfortunately, some people may lose their jobs as organizations invest in AI and need to account for and justify the investment, but the reality will set in that success with AI still requires skilled workers. It will be important for tech workers to embrace AI and skill up on the technology to keep their skill set relevant for future opportunities.
How Can We Use AI for Enterprise IT?
At the recent EUC Unplugged conference, the great Jaymes Davis talked about generative AI and some of its uses for generating virtual desktops. He showed some of the awesome power of using various AI tools for achieving this task but also highlighted the need to prompt the AI with correct parameters. An example here would be asking AI to create you a PowerShell script to automate the packaging of an application. It may provide you a script that does successfully automate the packaging of an application. However, you will likely discover that the script is only about 90% complete because you did not ask it to include error handling at various points of the script. Also, the script may work for some applications, but unable to handle others you may encounter.
For example, you may ask it to develop a mobile application for displaying a conference schedule that allows attendees to build their own schedule using Java. It may return you a working mobile application, but the script may not include features like garbage collection to ensure it does not eat up excessive amounts of memory. Success with Gen AI requires explicit parameters passed to it by someone with the knowledge required to teach the Large Language Model how to do it properly.
![](https://www.numecent.com/wp-content/uploads/2024/06/Untitled-design-3-1024x1024.png)
AI has also been infamous so far in terms of generating images with incorrect text, people with missing limbs or too many fingers, and many other anomalies (seriously, zoom in on the birds in this image).
The work AI generates can be useful building blocks for seasoned developers and the images AI generates can be modified by talented graphics designers. This could speed up the turnaround time for a project by automating much of the work, but it requires a human with the know-how and experience to use effectively. It could improve the work lives of many people but is unlikely to replace everyone. Large Language Models driving GenAI are also only as good as the datasets they are fed. This data is generated by humans. All of us train these models and as industries, technology, and everything continues to advance. The LLMs will need to be continuously trained and fed data.
If you are beginning your journey with AI, I suggest you quickly learn how to embrace the technology in a sensible manner. Create policies around the use of AI services in your organization to educate employees to not input company data into public AI services like ChatGPT. Inform those using AI services that they must view the results with a critical eye as some of the information may be incorrect. Incorporating this type of information in existing mandatory employee training is critical.
Explore possibilities of introducing solutions like Copilot for GitHub to assist your technical teams and ChatGPT, Baird, Copilot, or other services to assist employees where it makes sense. For example, to fine-tune marketing content, look at your automation solutions to see if AI services can provide greater efficiencies. You must always assess the true value that products which claim to feature AI bring. There is a lot of marketing FUD out there, so do your research. Do not be swayed simply with the use of buzz words. Assess the products and features on their merit and the value they bring to your organization.
I am excited by what AI and Machine Learning can bring to our industry. I hope you are too.
Subscribe to the Office of the Technologist
The Office of the Technologist evaluates the latest methodologies, Windows technologies, and industry trends to determine the best approach for delivering and optimizing modern, enterprise-grade IT operations. Fill out the form below to be among the first to hear about the latest content.