The COVID-19 pandemic has accelerated machine learning (ML) adoption in many areas, resulting in firms increasing their ML investment and implementation efforts. How can emerging markets like Latin America take the opportunity to embrace and adopt artificial intelligence (AI) and ML models more quickly?
- The 2020 Refinitiv machine learning survey confirms AI/ML adoption continues to grow globally, with North America leading adoption rates.
- This represents a huge growth opportunity for enterprises globally, yet the report findings suggest that Latin America is expected to only see gains of 5 percent due to lower ML adoption rates.
- Despite technology and investment gaps compared to North America, Latin American enterprises can still address the challenges with better resources.
For more data-driven insights in your Inbox, subscribe to the Refinitiv Perspectives weekly newsletter.
The 2020 Refinitiv machine learning survey confirms that ML adoption continues to grow globally, with North America leading adoption rates. Seventy-two percent of firms now say ML is a core component of their business strategy.
In many areas, the COVID-19 pandemic has accelerated ML adoption. While COVID-19 caused major disruption and upset ML models — resulting in firms increasing their ML investment and implementation efforts in North America — emerging markets like Latin America have the opportunity to embrace and adopt AI and ML models more quickly.
Read the 2020 Refinitiv machine learning (ML) survey
Growth opportunities powered by ML
According to a new worldwide financial assessment carried out during the COVID-19 pandemic, the future scale and impact of AI technologies will surpass the current trend, with the adoption rate reaching a new level.
For Latin American enterprises, this represents a huge growth opportunity: An Inter-American Development Bank (IDB) report estimates that AI influence could add 14 percent to GDP growth globally by 2030. Although all geographical regions are poised to experience AI’s economic benefits, Latin America is expected to only see gains of 5 percent due to lower ML adoption rates.
However, that’s not set in stone. Latin American enterprises can grow more quickly and outperform competitors by accelerating AI adoption.
The Refinitiv survey found that 34 percent of AI adoption by financial enterprises in the Americas is driven by the desire to stay ahead of the competition.
Latin American companies tend to lag behind their North American counterparts in AI adoption, which leaves significant gains open to first-movers and early adopters. Businesses increasingly rely on ML as a critical tool to generate value, manage risk, and achieve competitive advantage.
Challenges and solutions for Latin America
While Latin American finance enterprises tend to follow the lead of their global counterparts when adopting new technology like AI/ML, they face some unique challenges. Fortunately, these challenges are becoming easier to solve.
Despite the technology and investment gap when compared with North American counterparts, Latin American countries can choose to optimize opportunity and address the challenges with better resources. Countries in the region should understand the ongoing technological changes and develop a national strategy to seize the opportunities.
The language barrier
Natural language processing (NLP) is a subset of AI, enabling firms to extract and gain insight from vast amounts of language-based data. However, NLP requires access to pre-trained models in specific human languages.
For Latin American businesses, benefiting from natural language is complicated by the scarcity of appropriate AI models. While there’s often plenty of data, pre-trained ML models, and application programming interfaces (APIs) available in English, it has traditionally been difficult for Latin American companies to acquire or build equivalent training datasets in Spanish and Portuguese.
However, this is changing. With the increasing adoption of ML by Latin American firms, relevant data sets are quickly becoming more available.
Education, training, and talent acquisition
Education and training are also a challenge to ML adoption among Latin American businesses. Improving national education systems and increasing tertiary education is mandatory for preparing the next generation, but not sufficient. The current generation of South American workers also needs to adapt to the AI economy.
It has often been difficult to acquire talent with the right mix of data science, ML, and business skills in Latin America. As mentioned in the IDB report, according to the International Labour Organization, only about 20 percent of South American jobs require high-level skills, compared with over 40 percent in the European Union and the United States.
Automation and AI change will potentially redefine the workplace landscape entirely, eliminating job posts and motivating workers to learn new technologies and seek positions more aligned with digital transformation. However, filling these job openings with people who have the technical skills required to design and maintain AI systems isn’t easy.
Latin America has considerable work to do in this area, but Latin American firms are already actively working with universities to develop a bigger talent pool.
Coordinated initiatives by private and public institutions, startups, and financial groups have been building an ecosystem to foster collaboration and co-innovation. Universities are expected to generate quality graduates for startups. Financial institutions then provide funding to develop products and services.
North American companies often poach top Latin American talent. The good news is this validates the quality of Latin American ML professionals. They’re being headhunted because their skills and training are world-class.
Latin American firms are overcoming this brain drain by providing compelling work environments and interesting problems to solve, giving top talent great reasons to pursue opportunities close to home. To attract and retain talent, research organizations should identify economic opportunity areas and the potential AI and automation types to address them.
The impact of COVID-19
In many ways, the 2020 pandemic caused a significant reset for companies around the world, and Latin America is no exception.
Enterprises were forced to rapidly re-think policies and processes — and ML has grown increasingly important. Among the firms surveyed about how their investment strategy changed as a result of the pandemic, 40 percent are accelerating investments in ML due to COVID-19, 51 percent are maintaining the same level of investment, and only 8 percent are decreasing investment.
This information is especially important for Latin American companies because they tend to be more conservative than their North American counterparts.
Enterprises cutting ML investment to ride out the economic storm may find themselves far behind competitors who are using the pandemic as an opportunity to double down on AI.
Data scientists using ML and alternative data — or ‘alt data’ — apply analytics to reveal additional insights not previously available from traditional financial and business sources. Alt data includes a variety of new data sources such as AI-powered image recognition. Many investors believe alt data is as essential as fundamental data for their financial analysis and insights.
Alt data is now critical for spotting warning signals earlier and more accurately. It is no longer an ‘alternative’ because 97 percent of firms use it for ML. It’s an invaluable tool for gaining an edge as the pandemic forces companies to reset their strategies and rebuild their models.
What’s next for Latin American enterprises?
Latin American enterprises face unique challenges in adopting AI and ML. However, ML is also critical to their continued growth. Refinitiv helps companies access the deep data they need to gain a competitive advantage.
Visit Refinitiv Labs to discover Refinitiv’s exciting new ML projects. Explore data exploration tool, a solution that gives data scientists, quants and developers free, easy and intuitive access to sample Refinitiv data sets and notebooks.
To learn more about the ML trends we’ve discussed, download your copy of the 2020 machine learning survey report