Artificial Intelligence (AI), the development of machine systems that perform intelligent tasks, is the new buzzword of the media, as it promises to change the business world and everyday life.
“The hype around generative AI like ChatGPT is quite well-earned, especially for improving our digital workspace,” Kevin Gordon, VP of AI Technologies at NexOptic, an AI imaging solutions company, told International Business Times.
“These generative AI services can accelerate workflows by creating first drafts of documents, procedures, and policies. In addition, they can be used as so-called ‘oracles’ that help users summarize vast data stores. Currently, search engine results return the highest-ranked pages, and the user sifts through the pages. But a tool like ChatGPT can analyze those pages and provide a summary.”
Dr. Uri Yerushalmi, Chief AI and Co-Founder of Fetcherr, provides additional insight into AI systems’ capabilities and promises. “The recent progress in language models (as seen in GPT models) and model explainability techniques open the door for models to be much more white-box and transparent for their end users,” Dr. Uri Yerushalmi told IBT.
“AI products that can explain themselves are likely to automate more and more decision making in enterprises and finally breach the entry barriers previously placed by sophisticated but black box types of technologies that failed to do so. “
While the buzz surrounding intelligent search engines like GPT (Generative Pre-Trained Transformer) and chatbots is new, the hype for the promises they hold to change the world is hardly new.
Five years ago, AI was the buzzword of the media world, but it was about self-driving cars—cars driven by machines. They also promised to change the world, making driving safer and more efficient.
Nowadays, the promise behind self-driving cars remains, with the hype giving away to reality: self-driving cars may be years or even decades away from taking over the streets and the highways, as they have yet to overcome several technological and social barriers.
Daniel Reichman, Ph.D., an engineer who has worked closely on AI programs at Duke University, attributes this gap between hype and reality to personification.
“The concept of AI is that there exists a path to mathematically/programmatically mimic a human (and even a super-human) thought process,” he told to IBT. “So, at every corner, it is tempting to say: because the AI is capable of performing a certain task, it must also be capable of performing another somewhat similar task. Or, worse yet, because it can perform a task (e.g., recognizing a concept) that has a real-world understanding of this concept.”
Reichman believes the world is far more complicated than that.
“A real world-understanding of a concept (e.g., a lion) includes more than just recognizing it in pictures,” he said. “But also knowing the context of lions, e.g., where lions appear, their place in the food chain (and what that even means), why a lion would be found outside its natural habitat (what is a natural habitat). Instead, an AI is typically provided with many image examples of lions and is told to find lions — with no additional context provided.”
Yerushalmi provides further insight into the gap between the promise that surrounds AI systems and the prospects of becoming a reality and gaining broad acceptance in the business and everyday world.
“One of the obstacles that block ML & AI systems from being deployed in enterprises is the existence of legacy systems in the enterprises that could not co-exist with new systems,” he explained. “In 2023, we will see more and more cases where the ML/AI systems that use advanced techniques, including AI-based models, not only for their main purpose of serving the enterprise but also for having seamless integration into the enterprise, in addition to fast and yet gradual & risk-less onboarding.”
Yerushalmi believes that the economic pressure many enterprises face will force them to overcome this obstacle. “The combination of the 2023 financial circumstances and the AI golden age would result in a higher tendency to replace old systems, structures, and practices with new ones that generate higher income using fewer expenses,” he added.
Editor’s note: This is the first in a series of articles on how AI is changing the business world.
