Information overload caused by the constant data collection of digital crawlers and other algorithms helps obtain answers to questions that usually have an extremely short expiration time. Nevertheless, some commonalities do not change. The latest novelty captures all the attention until it is replaced by another event that seizes society’s focus. In this way, one day, the attention on politics might center on a president with dictatorial ambitions claiming to be above the constitution. Then, the day after, the problem is forgotten as a famous actress is cast for the main role in another superhero movie or because it’s more important to know whether a photo of an English princess has been retouched or not. This rapid shift in societal attention due to information overload is a cause for concern. It’s a call for our collective engagement, a responsibility we all share.
The oversupply of content seems to have hijacked minds, turning wars into geography lessons, where people can pinpoint the belligerent countries and the current war zone on a map.
The world of technology is not much different; we often encounter a technological term that seems responsible for humanity’s complete transformation. It’s crucial to understand the societal implications of these technologies. A technology that will allow us to leap forward several decades, promising to take us to Wakanda, the digital utopia that seems repressed in our souls. Thus, at one point, blockchain emerged as the technology that would drive a paradigm shift in digital transactions and security systems.
The intense discussion surrounding this technology gradually faded into the background as the promises of 5G began to materialize. If the most radical evangelists were to be believed, this new mobile generation would force humanity to discard the Bible, the Torah, and the Quran, as none of these religions could emulate the benefits of IMT-2020. According to its prophets, we would finally have a technology that would allow the massification of remote surgical procedures and end the existing multiple digital divides. Not to abide by its power is being warned by doomsayers predicting a technical apocalypse for any country delaying its 5G launch.
As the 5G topic matured with a steady growth of commercial networks during its first two years, a new word became popular in our lexicon: the metaverse. First used by novelist Neal Stephenson in his 1992 work “Snow Crash,” this word is now being driven by a digital company that wants to create parallel universes in cyberspace, going beyond what “Second Life” achieved or what can currently be seen with “Roblox” and other games.
The hidden irony is that when a writer creates a word, the world hardly reacts, but when a giant of the digital world uses it, we suddenly can’t live without reaching the metaverse. It became the raison d’être for many entities that began allocating funds to it. Curiously, after several years of being on the road towards the metaverse, a specific definition of the metaverse can’t be provided. The old premise is not knowing how to explain it but being able to identify it when facing it. Thus far, no one has seen it, and the metaverse continues to be a billion-dollar abstract concept.
Unfortunately, the launch of ChatGPT in November 2022 redirected the global interest in the metaverse with that of generative artificial intelligence in its text form. Siri, Alexa, and the other predictive AI audio versions were suddenly forgotten. ChatGPT and the ever-growing selection of generative AI competitors gave many evangelists the excuse to claim that we can leap into the future with the “arrival” of AI. However, for this maxim to hold, all existing versions of AI must be forgotten or, at least, ignored.
The reverence for artificial intelligence continues and has even intensified with the announcement of the European Union’s approval of a law to promote the development and definition of Artificial Intelligence. The enthusiasm surrounding artificial intelligence has received its holy scriptures; now, it’s simply a matter of spreading the good news to everyone. Now, we will finally emerge from poverty, genocides will end, and humanity will be able to rest as machines will relieve us of many tasks. AI provides a plethora of positive externalities!
Everything written up to this point highlights two significant problems that must be addressed to continue the anticipated drive for economic growth and social development promised by including a broader role for AI in ICT. There are two inevitable truths: no technology alone will solve all of society’s problems. Second, technology does not exist in a self-sufficient bubble; it depends on many factors to be implemented and utilized.
All the technologies mentioned earlier—blockchain, the metaverse, 5G, and artificial intelligence—require investment in tangible infrastructure that can create the telecommunications network necessary for high-speed data transmission. Additionally, 5G, a wireless technology, needs radio spectrum to provide its services. The investment to deploy and use these technologies successfully will depend on the existing telecommunications infrastructure and the availability and quality of civil infrastructure.
Of course, all that information would still be insufficient to determine the levels of investment; it is necessary to verify the existing legal framework, especially concerning permits and tax issues, the number of experts being trained by universities, import and export tariffs, demographics and topography, user profiles, and terminal costs, among other factors.
Talking about blockchain, the metaverse, or artificial intelligence is a waste of time unless these elements are reviewed to achieve connectivity and accessibility.
Now, imagine a world where all these connectivity problems have been overcome. Let’s focus on artificial intelligence. These systems are only as good as the information they receive. Their inputs are data, but if the input they receive is incomplete, the results that can be expected from this system will be far from optimal. The big problem in Latin America is the lack of concrete data that can be used and the lack of digital inputs to feed these systems compared to what is available in economies like China, the United States, Japan, and the United Kingdom.
The coverage that the European Union’s approval of its Artificial Intelligence Law has received paradoxically highlights the significant problem these systems face when they reach Latin America and other developing regions: misinformation and the availability of erroneous content or data.
For example, it is widely mentioned that the European Union’s law is the first to be approved globally. However, this information is inaccurate. Looking at Latin America, we can see that in Peru, 2023 Law 31814 was enacted, also known as the “Law that promotes the use of Artificial Intelligence for the economic and social development of the country.” The promulgation of this law in Peru led the global consultancy Access Partnership to state in July 2023 that “the Peruvian legislation sends a positive message to other regulators and legislators who have recently discussed similar initiatives, such as in the case of Colombia, Mexico, and Brazil.”
Peru’s Law 31814 contributes to the understanding of artificial intelligence by including a definition divided into four segments:
- Emerging general-purpose technology: Technology with the potential to improve people’s well-being can contribute to positive, sustainable global economic activity, increase innovation and productivity, and help address key global challenges.
- Artificial intelligence-based system: An electromechanical system that can, for a range of objectives defined by humans, make predictions, recommendations, or decisions, influencing real or virtual environments. It is designed to operate with different levels of autonomy.
- Emerging Digital technologies: Can generate innovative solutions such as robotics, analytics, artificial intelligence, cognitive technologies, nanotechnology, the Internet of Things (IoT), and similar technologies, forming Industry 4.0. This industry combines advanced production and operation techniques with technology, impacting the digital ecosystem, organizations, and people.
- Algorithm: A sequence of instructions and ordered and finite sets of steps to solve a problem or decide.
Similarly, a study published in January 2024 by the IAPP (International Association of Privacy Professionals) described in its “Public Policy and Global AI Law Tracker” report the initiatives to regulate artificial intelligence across 24 different jurisdictions, including those that had already incorporated laws regulating the technology, such as China or the United States. It also mentioned government initiatives related to artificial intelligence, such as creating a Ministry of Artificial Intelligence in the United Arab Emirates in 2017.
The great importance of the law approved by the European Union lies in the fact that it incorporates a law on artificial intelligence into the regulations of its 27 member states. For the rest of the world, it is relevant because it is the most comprehensive law on this subject and will serve as a model for many other jurisdictions, either to improve existing laws on the topic, as has already been announced in Peru or to fill a regulatory gap.
The Artificial Intelligence Law approved by the European Parliament has as its main objective:
- “Address risks specifically created by AI applications
- prohibit AI practices that pose unacceptable risks
- determine a list of high-risk applications
- set clear requirements for AI systems for high-risk applications
- define specific obligations deployers and providers of high-risk AI applications
- require a conformity assessment before a given AI system is put into service or placed on the market
- put enforcement in place after a given AI system is placed into the market
- establish a governance structure at European and national level.”
In other words, it takes an approach focused on risk minimization, determining the legality of its use depending on the harm it may cause to society. This is achieved by segmenting usage into four risk levels—unacceptable, high, medium, and minimal—providing great flexibility to the approved regulation.
However, we cannot deceive ourselves into thinking that the approval of an AI regulation will be sufficient to establish the parameters for its appropriate use in the long term. The very evolution of artificial intelligence indicates that in the coming years, we will move from the current “narrow AI” systems and begin to enter the world of “general AI,” which precedes “super AI,” events that, according to Ray Kurzweil, should occur in 2029 and 2045. The nature of these advancements raises questions about the effectiveness of the recently approved framework in achieving monitoring, implementation, and compliance with these rules for advanced systems.
Suppose the creation of a new language by two chatbots of a narrow AI system from Facebook sparked the worst nightmares, albeit exaggerated, of an apocalyptic future. What safeguards can be placed against systems that equal or surpass human intelligence? This debate continues and will become more heated as it opens up political, technological, philosophical, and ethical discussions.
The need for AI legislation is increasingly evident as debates arise over intellectual property, such as authorship of works or protection under fair use of protected works. Beyond the legal realm, artificial intelligence in Latin America requires networks capable of transmitting data at appropriate speeds, and it urgently needs to develop the human capital capable of working in this segment of technology. This implies significant investment in telecommunications infrastructure and university educational institutions.
This does not prevent us from seeing some amendments to the new European AI regulation in less than a decade and a half, as the global trend has been that less and less time passes between reforms to legal frameworks focused on technology. For example, there was a 23-year gap between the European Parliament’s approval of the “Data Protection Directive” in 1995 and its replacement by the more comprehensive “General Data Protection Regulation (GDPR).”
It is important to highlight that although having a comprehensive but flexible regulatory framework for AI is imperative, numerous other factors, including human capital development, will determine how quickly its different variants are adopted in the productive segments of the economy. Many lessons from the EU’s AI Law are not new for Latin America. Many of the concepts regulated by this norm have been discussed in numerous forums throughout the region, including four important events held in 2015 in Puerto Rico, 2023 in Uruguay, 2024 in Chile, and more recently, in Colombia:
- 2015 – San Juan, Puerto Rico: The Future of Life Institute held its conference, “The Future of AI: Opportunities and Challenges,” with the participation of some of the world’s leading experts on AI. The conference produced the document “Research Priorities for Robust and Beneficial Artificial Intelligence,” which listed as the main priorities for AI the optimization of its economic impact and research on its diversity of legal implications across industrial sectors, research in computing, control, and security. Participants in the forum agreed that AI’s impact on human society is unstoppable. Thus, they concluded that “it is the duty of AI researchers to ensure that the future impact is beneficial.”
- 2023—Montevideo, Uruguay: Organized by Khipu Event Series, the 2023 edition of the “Latin American Meeting on Artificial Intelligence” concluded with the adherence by more than 100 stakeholders attending the meeting of the “Montevideo Declaration on Artificial Intelligence and its Impact in Latin America.” The document calls for an anthropocentric approach to AI so it can generate well-being for humankind by strengthening the protection of human rights, respect for cultural and religious differences, minimizing its environmental impact, and the need for Latin Americans to be active participants during “the research and development [process], and not only as mere producers of raw data or handwritten notes with low added value.”
- 2023 – Santiago, Chile: The “1st Ministerial and High-Level Authorities Summit of Latin America and the Caribbean,” held in Chile in 2023 and summarized in the Declaration of Santiago, called for the creation of an Intergovernmental Council on Artificial Intelligence for Latin America and the Caribbean working group that complies with UNESCO’s recommendation on ethics and AI. The working group would invite the InterAmerican Development Bank and CAF to join the task force during its first two years. The 2nd Ministerial and High-Level Authorities Summit of Latin America and the Caribbean is scheduled for October 2024 in Montevideo, Uruguay.
- 2024—Cartagena de Indias, Colombia: The Colombian government organized “The Latin American and Caribbean Ministerial Summit on Artificial Intelligence ColombIA.” The event reunited representatives from 17 governments from Latin America and the Caribbean to discuss the importance of approaching AI ethically and responsibly. Its main objective was adopting the “’Cartagena de Indias Declaration for Governance, the Construction of Artificial Intelligence (AI) Ecosystems, and the Promotion of AI Education in an Ethical and Responsible Manner in Latin America and the Caribbean.” The declaration has three main pillars, which signatories accorded to cooperate and coordinate their efforts to promote the best practices in AI governance, digital education about AI, and enabling digital ecosystems that leverage AI.
We can infer several lessons from all these international efforts to guarantee the ethical use of AI technologies, their adoption by Latin American governments and companies, and the modernization of curriculums at all levels of education to highlight the increasing importance of digital technologies. First, governments throughout the region acknowledge the importance of AI and are conscious of the impending normative challenge that all its different versions will create across all industries. AI will exacerbate the challenge of regulating social media players as it will impact all aspects of life and productivity.
Second, despite all the abovementioned efforts, data availability will drive or hinder artificial intelligence in Latin America. It is a mistake to think that deriving it from people’s interactions with a social network or mobile device will be enough. That is simply one part of the equation; significant work, investment, and education are needed to build the rest. Third, Latin America cannot aspire to be a global AI leader until its governments allocate a budget for research and development. Latin American countries cannot aspire to lead the global AI conversation if their countries do not invest in the development of new AI patents, services, and products. Without an increase in the number of AI patents, AI academic publications, and well-funded technology centers for AI, Latin America and the Caribbean will continue to be reactive players in the AI landscape. According to Stanford University’s “AI Index Report 2024,” Latin America is almost non-present in AI’s productive cycle.
We should continue the important discussion on ethical and normative aspects of AI. Still, we cannot expect Latin America to be a leader and have a global impact unless local research and development focused on addressing local concerns increases in the region. Avoiding investing in AI will result in the area continuing its role as a follower of global technological trends. AI is becoming a double-edged sword that could increase efficiency and productivity while worsening Latin America and the Caribbean’s multiple digital divides that so much pain creates throughout the region.
