IT Pro Today (2023-01-04)
A confluence of many factors – powerful computing in small form factors, edge computing, the integration of IT and operational technology (OT), 5G, and even the COVID-19 pandemic – has buoyed the adoption of artificial intelligence in a variety of industries. Valued at $93.5 billion in 2021, research firm Market View Research predicts the global adoption of AI will grow at an astounding 38% compounded annual growth rate until 2030.
So where does AI go from here? The experts shared six AI predictions for 2023.
Related: 5 Ways to Prevent AI Bias
1. Generative AI will continue to garner attention, said Mike Krause, AI solutions director at Beyond Limits, an enterprise AI software firm. Generative models, like the digital image generator DALL-E, analyze data and interpolate to create something brand new.
But generative AI models are not just good at creating digital images like DALL-E does. They are being used to discover new materials for battery design, carbon capture, and loads of other innovations, Krause said, who predicts that generative models will reach new heights in 2023. For example, expect developments in the healthcare space for vaccine modeling, drug discovery, and even personalized medicine supported by training data generated from electronic medical records, Krause said.
2. AI will become less of a black box, said Lee Howells, head of artificial intelligence at professional services firm PA Consulting. Howell predicts 2023 will see more organizations voluntarily publicizing their AI principles and outlining their processes.
“There will be greater use of ‘explainable AI’ over black box models in areas that directly affect individuals,” Howell said. “Organizations with published AI principles and demonstrably ethical use of AI and data will see greater acceptance by the public of the use of their data.”
“As more and more AI [systems] are being deployed across various sectors, regulators are keen to ensure that all AI models are behaving the way they should, free from any bias and discrimination,” noted Lian Jye Su, AI and ML research director at ABI Research. He argued that while explainable AI will make the process more transparent, it requires several market developments to happen, including a supportive model development and deployment infrastructure that can show the relationship between the inputs and outputs and processing layers, the introduction of AI models that are explainable by default, and clear regulatory guidelines and principles.