2 days ago
- Business
- Business Journals
AI revolution: Navigating the promises and perils of a new era
At electronic health records innovator ModMed, its proprietary artificial intelligence application, ModMed Scribe, has been trained on millions of patient encounters. It is improving how doctors interact with, diagnose, and care for patients to improve office workflow.
JM Family Enterprises use AI-powered tools to help create and test new applications. This is improving the quality of software development and reducing the time required to produce requirements and test cases by approximately 30%.
A contact center platform with AI capabilities at BankUnited is exploring how sentiment analysis can help bankers and customers find the solutions that best fit a given situation.
Miami Dade College, in collaboration with Microsoft, is applying the technology company's 365 Copilot AI-powered productivity tool atop its suite of applications to drive staff and faculty efficiency. Since implementation, nearly one year ago, adoption is over 90%, which is impressive, understanding that it's a brand-new tool, according to Antonio Delgado, the school's VP of Innovation and Technology Partnerships. This is simply some novel large language (LLM) model application; it's facilitating efficiencies across the school's technology suite.
'The results that we see even before getting to a full year of implementation is an enormous number of hours saved by leveraging the tool,' Delgado said. 'We see greater satisfaction. We see the productivity. And at the end, the outcomes with our students are better communication and better sense of belonging.'
AI's promises and potential perils were explored during the recent AI Revolution Roundtable. Held at the Miami offices of the South Florida Business Journal and moderated by Andrew Duffell, president of the Research Park at Florida Atlantic University, leaders from some of the region's leading private and public sector employers discussed the challenges they're facing with artificial intelligence.
Deploying AI in the workplace
From the debut less than two years ago of generative AI applications like ChatGPT, companies today are turning to agentic AI. These systems can autonomously make decisions and act to achieve goals, often with minimal human intervention.
Already, AI in all its forms is delivering a 'noticeable shift from prioritizing large, transformative use cases to simply understanding how to best integrate AI into routine tasks,' said John Damalas, group VP and CTO with JM Family Enterprises, the Deerfield Beach-based automotive distributor and services company. AI is creating an evolving narrative that positions it as a support tool for human-led activities, rather than a total replacement for cognitive work.
JM Family's AI Center of Excellence has developed governance frameworks via its AI Council, a cross-functional team that promotes AI initiatives and ensures strategic alignment. It prioritizes responsible AI use with a 'formal process to review use cases for transparency and compliance with regulations, including monitoring proposed regulations in relevant jurisdictions,' Damalas said.
Julio Jogaib, CIO with BankUnited, shared that employees must be trained on AI's best uses to advance their organization. This includes creating governance protocols to create standards and guardrails for proper AI usage by the workforce and organization.
BankUnited's vendors are using AI in the applications they provide. Jogaib works to ensure the bank and its vendors 'have governance in place surrounding usage, data privacy, and other areas,' he said. It's especially important when client data is being used by vendors to train their systems.
'There is a lot of legal apparatus that needs to be in place and monitoring on what's being done by these platforms,' Jogaib said. 'They may say that they are not using our data. But you have to verify that.'
As a health care technology company, ModMed works to ensure the company and its partners have the rights to use data or documents in the development of its platform, from widgets, chatbots, agents or generative AI. Daniel Cane, the company's co-founder and co-CEO, represented ModMed on the roundtable.
'They might have an asterisk by it now and say, 'if you use this new AI agent, we're collecting information to understand how our agent works and better train it in the future,'' Cane said. 'Be careful when you're using third party tools to understand how they're going to use your data.'
Jogaib put it more fundamentally: 'You need to have people inside the company capable of thinking.'
Training next-gen users
Two cohorts of 'students' are emerging in the AI realm. One will be those traditional students in classrooms who are learning coursework to prepare them for future careers. Second are employees who are learning skills - including the use of AI - to serve their companies or future career opportunities.
Technology has changed education delivery. Think back to the pandemic, when classrooms from grade school to higher education embraced portable devices and remote learning. AI is continuing that evolution. It's 'personalizing the education journey,' Delgado said. It can identify in-demand jobs, and the skills, degrees, or certificates needed to pursue them.
The emergence of LLM applications, like ChatGPT, has students asking a question of the application, then copy/pasting the answer into their schoolwork. How do educators, or even employers, ensure students or workers actually learn and know the material?
'That's a big challenge that I see… the student is not doing the thinking anymore,' Delgado said. 'It's having the skills to embed AI in what you do and still add value, to defend the answer, to think about the abstract. We need to work on changing the way that academia is not only teaching but also assessing learning. Because if not, there is no learning happening.'
JM Family invests in workforce development and views AI as an enhancement, not a replacement, for human capability. A critical part of its AI Center of Excellence is giving associates the tools and training needed to learn about AI and understand how to leverage it within their work.
The roundtable's moderator, Andrew Duffell of Research Park at FAU, wondered how ModMed is encouraging employees to integrate their own knowledge with AI. Cane explained that the expectation is that employees are problem solvers and eager learners, and they arrive with a fundamental understanding of AI and its potential uses in the organization. Employees' natural curiosity must have them exploring how AI can be a tool to improve effectiveness, he emphasized.
'We are an employer that's using AI throughout our company, whether it's finance, accounting, marketing, software development,' Cane said. 'But we're also creating our AI to solve problems out in the wild.'
AI's misperceptions, challenges and threats
'The belief that AI will automatically enhance processes and products can be inaccurate,' Damalas said. 'Similar to the digitization trend of the late 1990s and early 2000s, integrating AI into a business process with considerable inefficiencies may only heighten those inefficiencies, unless the process or product is carefully assessed concerning the specific outcomes or goals that AI is supposed to achieve within its context.'
AI can create workplace challenges. According to Damalas, one involves managing change and promoting psychological safety for users as they adopt these technologies. He believes it is essential to address concerns that AI and other emerging technologies may replace jobs and instead encourage users to perceive them as enhancements to their skills and capabilities.
Another will be how industries, such as health care, banking and financial services or education, for example, work with policy makers to enhance future regulations around AI.
ModMed works with industry governing bodies to advocate for responsible AI applications and implementation. Internally, the company ensures its use of data complies with patient privacy laws, as well as emerging AI responsible-use and transparency guidelines.
A company like ModMed's unique place as a developer and user of AI also gives it a role to play as advisor to governing bodies.
'Now is the time to be working with policy makers,' Cane said. 'It's essential that we bring policymakers and governing bodies up to speed on how this content and how this technology can be used in a way that can help people, and try to prevent people from getting hurt by it.'
Where some may see regulators or governing bodies as 'the enemy,' Jogaib said the pace of industry innovation places developers and users in an important role as advisor.
'We call regulators our partners,' Jogaib said. 'We show them what they are doing, how we are doing it. We allow them to ask questions and through those questions they not only understand better how we are doing things, but they also take feedback. There must be this interaction, this exchange of experience.'
As with any technology, strong internal policies can help thwart misuse and external threats. Companies must bolster and update their cybersecurity apparatus, Jogaib affirmed.
Besides, as regulated industries, health care and banking, even education, face the prospect of government involvement - even suspension or revocation of licensure - should they not protect data or otherwise not meet regulations.
Dispelling misperceptions
Even as AI goes from headlines to frontline use in the workplace, developers and employers are still working to dispel fears and misperceptions that AI will replace workers.
In physician practices, Cane said that AI won't replace doctors. Rather, its ModMed Scribe platform is helping doctors focus on practicing medicine, as opposed to rote tasks and administration 'and all of the back office overhead on billing and collecting and working on payers and all the things that happen in the middle where a doctor wants more than anything just to focus on the human being in front of them, the patient,' Cane said.
'It's not uncommon for doctors to spend two or three hours after work in what we call, 'pajama time,' catching up on their notes,' he said. His AI platform takes notes during the visit, writes prescriptions, orders and reviews labs, and handles other documentation - even following up with the patient to ensure compliance.
From a more fundamental approach, Delgado noted that organizations need to educate their users - employees, customers, even students, in the case of educators - to ensure they're learning and using best practices.
'We need to really educate our constituents from that perspective about how to leverage it in a positive way, while still making sure they are open to including it as part of the journey,' he said.
To Delgado, as business and society embrace agentic AI, or any AI applications for that matter, it still requires human intervention to supervise and hold AI accountable for its outcomes. 'Hallucinations' or incorrect responses remain a serious concern whose prevention requires human oversight.
While future autonomous AI presents some dangers, the evolution brings tremendous promise.
The embrace of AI 'is not an isolated event. It's a team sport,' Delgado said. 'Everybody, every function, no matter what type of company it is, has a place in the AI game, even if they don't know that yet.'
