Карибские стартапы могут опередить крупные экономики в внедрении искусственного интеллекта

The Caribbean is not behind in the artificial intelligence (AI) race. In fact, its smaller economies and businesses may be better positioned than larger markets to adopt AI without the burden of legacy technology, systems and ways of working. That was the message from Mellissa Lezama, founder and chief executive officer of HR Horizon, during a Founder Institute Caribbean session, How to Lead Your Team Through AI and be an AI-Native Leader. Lezama said Caribbean founders had an opportunity to build AI into their businesses from the outset, rather than retrofit the technology into established systems. “The Caribbean region is not behind in the AI era. We’re actually structurally positioned to leapfrog it,” she said. Bigger, older economies, she said, were often burdened by legacy systems, contracts and workplace practices that could make digital transformation more difficult. Caribbean start-ups, by contrast, can design their companies, policies and workflows around AI from the beginning. “And even before our first hire,” Lezama said. Lezama, who has more than 25 years of HR experience across the United States, United Kingdom and Caribbean, has increasingly focused on helping start-ups and small and medium-sized enterprises build the HR frameworks needed to scale. She said AI was changing not only the tasks employees perform, but also how companies approach hiring, workplace policies, workflows and culture. The pace of change, she said, requires leaders to question established practices rather than simply adapt them to new technology. “Is your leadership style a choice or is it just a habit? Are you leading on purpose or are you just doing what’s familiar?” Lezama asked. She said AI, digital transformation and changing workforce expectations were reshaping leadership simultaneously, requiring business leaders to continually reassess how they operate. Learning from others Lezama pointed to mistakes already made by companies adopting AI, including recruitment algorithms that screened out candidates based on race, gender or disability; data privacy breaches involving customer information fed into AI tools without adequate consent or disclosure; and customer-service systems producing inaccurate or harmful responses. She also highlighted the risk of employee burnout as businesses use AI to increase productivity without reconsidering the demands being placed on workers. “Every single one of those instances was a very expensive, very public lesson that somebody already paid for. But we don’t have to make those same mistakes. We can build the guardrails in from day one and that’s a leapfrog advantage,” she said. Lezama said the conversation about AI and employment should go beyond fears that technology would simply eliminate jobs. She cited World Economic Forum projections that AI could displace 92 million roles globally by 2030, while 170 million new roles could emerge over the same period, representing a net gain of 78 million jobs. But the jobs being created will not necessarily be the same as those that disappear, nor will they require the same skills. For Caribbean businesses, she said, AI could allow very small teams to accomplish work that previously required significantly larger organisations. She referenced the concept of the “start-up of one”, in which an individual can use AI for research, drafting, customer support, operations and product validation. “For Caribbean founders, that is not just an interesting idea. It’s actually the whole game. You now don’t need a 50-person team to compete anymore. You just need the right AI tools and the willingness to learn them,” said Lezama. But AI adoption does not eliminate the need for human judgment. Instead, she said, businesses increasingly need AI and data literacy, critical thinking, adaptability and strong human skills. AI and data literacy, she stressed, do not necessarily mean coding. “It’s speaking to the ability to read AI’s output critically and know when it’s wrong,” she said. That requires employees and leaders to understand what data an AI tool has been trained on or fed, verify its answers against trusted sources, and question its output rather than automatically accepting it. Critical thinking becomes particularly important when decisions are nuanced or high-risk. “The hardest skill is judgment. It’s recognising when a decision is too nuanced, it’s too high stakes, it’s too dependent on context that AI simply would not have, would not know,” she said. Adaptability is equally important because AI tools are changing rapidly. Lezama warned businesses against becoming too attached to a particular platform, noting that products can change significantly through updates or become inadequate as a business grows. But she placed particular emphasis on human skills. “Empathy, communication, relationship-building, trust,” she said. “Once everyone has access to the same tools, the tools themselves stop being your competitive advantage. Because your competitors can buy the exact same AI that you are using. But what they can’t buy is how your team makes people feel,” Lezama said. She said customers would remember how they were treated when something went wrong, while employees would continue to value leadership and human relationships even as more technical and routine work became automated. Redesigning the workplace Lezama warned businesses against simply introducing an AI tool and expecting existing job descriptions and workflows to accommodate it. “AI adoption fails or backfires when businesses just drop a new tool onto old job descriptions and workflows and just hope for the best,” she said. Instead, business leaders should examine what remains of a role after AI absorbs its routine elements and determine whether job descriptions and performance expectations reflect that new reality. She also urged businesses to establish acceptable-use policies covering which AI tools employees can use, what information can be entered into those systems and when AI-generated work must be disclosed. Workflows should clearly identify where AI can make a first pass and where a human must review, approve or escalate an output, particularly where personal information, money, health, safety or legal exposure is involved. Lezama recommended that businesses review their AI policies and workflows quarterly because the tools being used today may not be the tools being used a year from now. She identified four key areas businesses should address: employment and engagement contracts, disciplinary and grievance policies, data protection policies and AI-use policies. Contracts, she said, should be flexible enough to reflect changing roles, particularly where employees work in AI-augmented positions. “The role that you hired someone for in January is not going to be the role that they’re doing in June, especially for AI-augmented, AI-supported roles. And so, your contracts should not pretend otherwise,” she said. For AI-use policies, businesses should clearly establish which tools are approved, who owns AI-assisted work, what information is confidential and what information must never be entered into an AI system. Businesses should not wait for a dispute or data breach before establishing those rules. “It can be a simple one-page version. And at the end of the day, having something in place, even though it’s simple, beats having nothing in place after something has already gone wrong,” she said. Caribbean advantage Lezama believes the region’s position as a later adopter of many technologies can itself become an advantage. She described what she calls a “regional trifecta” of agility, hindsight and networks. Caribbean businesses can often make decisions quickly because they do not have the layers of bureaucracy found in larger organisations. They also have the benefit of seeing what has worked and failed in larger markets before adopting similar technologies locally. The region’s diaspora connections, meanwhile, can provide access to international expertise and networks. “When we stack that agility, that hindsight and the networks all together, we are not playing catch-up. We’re actually playing a whole different game entirely,” she said. But technology alone will not deliver that advantage. Leaders also need emotional and cultural intelligence as their teams become increasingly distributed across countries and time zones. Lezama argued that Caribbean workers already possess a degree of cultural fluency that can help businesses manage diverse teams. “If you’ve ever adjusted how you speak on a call versus with family or if you’ve grown up around relatives who split their time between here in the Caribbean or New York, Toronto, Miami, or London, you’ve been practising cultural intelligence your whole life,” she said. “That instinct for reading a room and adjusting how you communicate, depending on who’s in it, is exactly the skill that lets us lead teams that are split across islands, diaspora hubs, and time zones without forcing everyone into a narrow way of working,” Lezama said. Sharing the gains Lezama also urged companies to ensure that AI-driven productivity gains do not simply translate into greater pressure on employees. “If we can produce so much more, how can we now share those gains with our employees?” she asked. Businesses, she suggested, should consider how increased efficiency could create more focused work time, reduce workloads or even contribute to shorter working weeks. “The truth is, if your AI tools are always on, the reality is that your team does not have to be,” Lezama said. She also stressed the importance of transparency when employees’ roles are changing. Employees, she said, increasingly want purpose, growth and transparency, as well as leaders who can acknowledge uncertainty rather than pretend to have every answer. “None of this requires any of us to have all the answers,” she said. “We’re leading through uncertainty. And that doesn’t mean that we need to be certain about or know everything or have all the answers. But it is about being honest with our team members about what we don’t know yet, while still moving forward with confidence in our own judgment,” Lezama said. Four safeguards Lezama identified four safeguards she considers particularly important for businesses adopting AI: data privacy, bias, burnout and judgment. On data privacy, she recommended that businesses clearly tell customers how their data is being used and examine the terms of service of AI tools to determine whether information is used to train models or shared with third parties. On bias, she warned that any AI system that screens, scores or makes decisions about people can produce discriminatory outcomes. “Anywhere where AI decides who gets what, bias can quietly creep in,” she said. She recommended auditing such systems before scaling them and repeating those audits regularly. For high-stakes decisions, she said, a human should remain in the loop. “Automation should speed up making a decision, but it shouldn’t be making it alone or replacing that final human overview and judgment,” she said.