AI (Artificial Intelligence) has been a key talking point in recent years, with speculation abounding about the technology’s potentially negative effects and whether the technology could really be implemented in many businesses. At the recent AI Summit in New York, sponsored by Informa, various sessions discussed how and where AI is being implemented, with the overall message being that the science has  learned from its growing pains and will become an integral element into companies’ business plans for 2023.

Natalia Nygren Modjeska, Director of AI Practice at Omdia, said during a session that in 2022, AI started to hit a critical mass. Quoting data from Omdia research, Modjeska said, “We see a quarter of organizations scaling AI, and 80% of surveyed respondents say they have reached a committed stage of AI deployment.”

Scaling Up Deployment

As companies ramp up their use of artificial Intelligence, they are learning it is better to tackle the various aspects of deployment in smaller steps and make sure they are achieving the desired outcomes, rather than trying to take giant steps that could backfire and send them back to square on.

One example has been AI data science expertise. A recent Omdia report said that constrained data science expertise will drive alternative AI solutions. The newness of AI has, not surprisingly, led to a dearth of personnel skilled in AI. To get around this, organizations are trying to resort to simpler development tools, according to Modjeska. She added this meant development tools which required little or no coding, software as a service (SaaS) solutions, off-the-shelf AI development packages, and embedded AI.

Data is also playing a key role in the mainstreaming of artificial Intelligence. Modjeska noted that companies are looking for innovative ways to manage data, including data catalogs and exchanges. In addition, there is now a trend away from “big data” to edge and smart data.

According to the Omdia report, having a clear plan to collect, curate, store, protect and process data will be a key foundation of any AI/machine learning strategy. More important, humans need to be at the heart of any strategy, as enterprises that do not follow this requirement will not be able to realize the full power of AI to guide their businesses.

Hardware for AI has also evolved as the science matures. “There has been major growth in AI silicon, with a more diverse ecosystem of chip types for AI.  Processors are trending from CPUs to GPUs,” Modjeska said.

The Omdia report noted that future AI silicon will build complex systems-on-chips and larger multi-chip systems through advanced packaging technologies.

Accountability and AI

One of the big concerns with artificial Intelligence has been taking responsibility when things go awry. With ethics and governance at center stage, Modjeska noted that companies need to determine who is responsible for ensuring that AI systems are ethical and free of harm, and who is held accountable when they aren’t. Companies need to invest in the proper tools, frameworks, services, and best practices to help achieve these goals, she added.

According to Omdia, 74% of the respondents surveyed in its report said regulations to prevent AI from having a damaging effect on data privacy is the highest priority for consumers. Another 46% of respondents called for regulations to foster transparency into how AI systems make decisions.

Modjeska added that organizations should also consider risk management at all stages of its AI ramp-up. “The potential costs of not doing this include loss of market share, brand equity, scrutiny, and potential actions from regulatory bodies. The lack of AI risk management costs organizations billions of dollars annually.”

The Good News

Despite a myriad of challenges, AI has proven to a solid return on investment for adopters so far, and should continue to do so in coming years. Omdia projects the AI software market to increase at a 31% CAGR (compound annual growth rate) through 2026, reaching $120 billion. Omdia added that companies will likely need to invest in in-house expertise to help drive their AI growth, in some cases supplementing that expertise with commercially deployed, point-to-point AI solutions.

Spencer Chin is a Senior Editor for Design News covering the electronics beat. He has many years of experience covering developments in components, semiconductors, subsystems, power, and other facets of electronics from both a business/supply-chain and technology perspective. He can be reached at [email protected].