Top CMSWire Contributors 2023: Spotlight on Frank Palermo

Frank Palermo’s passionate about AI, especially in how it impacts business. Palermo, in his insightful article “Executive Order on AI: A Needed Step or Kitchen-Sink AI Governance?” delves into the complexities and nuances of this very topic. As a digital experience expert as head of the Global Technical Solutions Group at Virtusa and with a keen eye for the intersection of technology and policy, Palermo expertly navigates the intricacies of President Biden’s recent executive order on AI.

His article, a blend of critical analysis and forward-thinking insight, not only examines the potential impacts of the executive order but also raises pertinent questions about its implementation and future implications. Palermo’s piece is more than a mere commentary; it’s a deep dive into the evolving landscape of AI governance, balancing between innovation and regulation. His expertise, shaped by years of experience in the field, comes through as he addresses the challenges and opportunities presented by AI in today’s world.

And this is what Frank does so well.

Frank, writing for CMSWire, leverages his profound understanding of digital technologies to offer readers a nuanced perspective on AI’s current state and its potential trajectory. Through his narrative, he skillfully presents a multi-dimensional view of AI governance, discussing its implications for businesses, governments, and the broader tech community. His articles are must-reads for anyone interested in the future of AI and its regulatory environment.

Q&A With Frank Palermo

AI Governance: Data, Science, and Policy Evolution

In light of your article on the executive order on AI, what are the key elements of effective AI governance, and how do you think governmental policies should evolve to keep pace with AI advancements?

The foundation of AI governance really starts with proper data governance and an understanding of the data lineage.

Data scientists are at the heart of developing AI models. While they provide deep experience in data relationships and logic, they may not have enough business acumen or context to understand the nuances of the data anomalies that can cause bias and model drift. The quality of predictions depends on the data you feed into the model. Tracking model drift is a key aspect of AI governance which requires keeping a constant pulse on the movement and context of your data. A proper AI governance framework should provide proper ongoing data transparency.

The government’s role in AI governance will continue to evolve as the technology matures. Currently, they are providing key principles and blueprints for an AI such as the AI Bill of Rights, which is useful in setting the foundational guidelines for responsible AI development while preserving open innovation and collaboration. The level of governmental oversight required is still not clear, but I expect something similar to the FTC or FCC will emerge for as a governing body for AI.

GenAI’s Future: Experimentation to Impact in Industries

Based on your insights into generative AI technologies, how do you see these technologies evolving in the near future, and what impact might they have on various industries?

2023 was the year of GenAI experimentation. The buzz around GenAI was a catalyst in boardrooms across the world to activate exploration. In some ways, it was a technical solution in search of problems and real business use cases.

We spent much of the year helping clients identify practical use cases and assist in the experimentation of GenAI solutions. Most of these centered around chatbots, service assistance, image classification, document validation and classification, marketing asset and campaign automation and automated information retrieval.

As we move into 2024, we’ll begin to see some of these experiments starting to gain momentum and begin some limited roll out. We are probably still far away from large-scale production class GenAI solutions.

The first wave of applications will likely center around worker productivity, with developer productivity being a big focus. The acceptance of these solutions is the realization that the worker of the future is not actually competing with an AI system for their job, but rather another worker who is being augmented with AI enabling higher levels of productivity.

As an example, companies like Microsoft have launched their Co-Pilot AI assistants to aid users in their productivity. It works across the product line with productivity applications such as Word and Excel. However, it gets really interesting for developers as it has the potential to drive significant productivity increases — currently being estimated up to 30%. This would have a huge impact for businesses as they continue to figure out ways to do more with less.

AI Power Struggle: OpenAI Drama to Tech Evolution

From your perspective, what factors are contributing to the current jockeying for dominance in the field of generative AI, and how might this shape the future of the technology?

Shockwaves went through the AI community at the end of November as OpenAI’s board mysteriously ousted CEO Sam Altman without providing any transparency on the cause. The drama that ensued in the following week was nothing Silicon Valley had ever seen before. In a shrewd move, Satya Nadella quickly appointed Sam Altman as head of a newly formed AI division to protect their AI interests. However, a petition of over 700 OpenAI employees threatening to leave if Altman was not re-appointed made the board reverse its decision and reappoint Altman as CEO. In the blink of an eye, OpenAI’s current AI dominance almost vanished. 

OpenAI created a call to action across the big tech companies to accelerate their strategies around AI. Meta, Google, X (formerly Twitter) have all responded with significant GenAI platforms. Google’s Gemini was launched this month and is rumored to outperform GPT 3.5 and possibly even GPT 4. Grok was launched by X and has direct access to X’s data repository making the responses more natural and conversational. Meta released its Llama 2 powered assistant called Meta AI.

While OpenAI may still be leading the pack in terms of user base and capability, its clear 2024 will be a year in which we see these gaps closed and possibly see convergence in terms of overall model performance.

Similar to the evolution of cloud services, GenAI capabilities will also become part of the computing infrastructure. They will evolve to be as ubiquitous and virtualized as cloud based storage and computing is today. In the future applications will be able to rapidly embed GenAI capability, making all applications more intelligent. 

GenAI’s History Shapes Its Future Potential

How does understanding the history of generative AI help in comprehending its current and future potential?

There is an old saying that history repeats itself. Many people don’t have an appreciation for the long eight-plus decade journey that AI has been through with many ups and downs along the way. Periods of tremendous progress were frequently followed by disillusionment and hyped expectations (a phrase coined as the “AI Winter”).