If AI was a hit new toy this year, 2024 will cement it as a tool.
AI, or artificial intelligence, grabbed the attention of every industry in 2023, creating hope for easier, more streamlined work processes and stoking fear that the technology could advance enough to replace employees.
While the technology world wrapped itself around the possibilities of AI this year, the business world is getting ready to put it to use and test just how much it will change the world beyond cheating on college essays.
Business leaders, from the airlines to commercial real estate, have been quick to tout the potentially game-changing uses of AI. A Southern Methodist University student created an app that uses AI to generate study guides based on lectures. Government leaders are pushing to create guidelines for how and when AI should be used.
Regardless of how you feel about it, AI isn’t going anywhere in 2024; Experts say it’ll only become more commonplace for companies to implement a number of AI programs to use in day-to-day business.
“Next year we will see across every sector, in almost every enterprise, some kind of an AI assistant at the employee level, whether it’s something as simple as drafting an email or putting together the first draft of a PowerPoint,” said Beena Ammanath, executive director of the Deloitte AI Institute.
The last 12 months have been particularly fruitful for generative AI, or programs that can create text, pictures and other media. Once hard to access and resigned to computer labs, AI is more democratized than ever. It’s a phenomenon perhaps best exemplified by the rise of the free chatbot program ChatGPT, launched in late 2022, which transitioned from a novel AI system to a part of the common lexicon.
The AI umbrella encompasses tools that range from large language models like ChatGPT to programs that direct autonomous vehicles. A type of AI that’s useful to one industry might not be useful to another, but nearly all sectors are looking to find the niche AI can fill in their own work. Around 94% of business leaders said AI is critical to success in the next five years, according to a Deloitte report.
In health care, a model could take existing patient data to predict how a disease will progress or how likely someone is to receive a certain diagnosis. In the public services sector, machine learning algorithms could evaluate the risk of housing or food insecurity.
Developments in language processing are particularly exciting and contain the potential to change how people from across the globe communicate, Ammanath said. People speaking different languages could have their words translated in real-time in their own voice in the coming years.
“It’s not too far that that communication could happen,” Ammanath said. “It just makes it much more seamless and human, in a way.”
With every technology comes risks, something AI has no shortage of. General Motors recalled its Cruise autonomous vehicles in November after one of them dragged a pedestrian to the side of a San Francisco street. Tesla issued its largest recall ever this month after a regulator determined the company’s driver-assistance system didn’t effectively guard against misuse. A number of news organizations have tried ChatGPT with mixed results.
AI can also “hallucinate,” or generate incorrect information, if trained on a dataset that’s unrepresentative or limited. People can’t trust with full confidence that an AI model will work perfectly.
“Accuracy cannot be guaranteed,” said Gopal Gupta, co-director of the Center for Applied AI and Machine Learning at the University of Texas at Dallas. “The next year will be focused on how to build applications that are reliable, that you can actually put in front of customers.”
As for the workforce, AI is unlikely to cause the catastrophic disruptions to jobs that some business leaders have predicted. Manufacturing-based industries like aerospace, automotives and electronics are less likely to be impacted than industries like banking, pharmaceuticals and education because of generative AI’s strengths in language activities over physical labor, according to a McKinsey & Co. report.
The need for humans to train and check AI models is also a reassurance that the technology won’t render workers obsolete, although employees will have to adapt to working alongside AI. Such tools are only useful if companies train their workforce on how to use them efficiently.
“We’ve truly entered that era where you need to be a lifelong learner. These technologies are going to keep coming at us faster,” Ammanath said. “It is no longer about just doing the once-a-year or when-you-join training and then you’re done.”
