The human brain now competes with hundreds of automated minds due to increased adoption of artificial intelligence (AI). In the case of generative AI, 55% of companies in Argentina, Colombia, Chile and Peru use it at basic levels.
The artificial intelligence (AI) boom began a year and a half ago and today, its accelerated expansion has exceeded initial projections that anticipated its adoption during a six-year process. Nowadays, this is done in six months. Currently, a person's brain works and competes with hundreds of AI brains, explained David Tolosana, Chief Digital Officer of Microsoft during Telefónica's Hispam Digital Forum in Peru.
“We will have several one-person unicorns in a couple of years,” he said.
But how much is hype and how much is real?
In the United States, there are those who hire companies to analyze embryos and, using AI, define which one is less likely to inherit diseases. There are also others that, by just using a photo, can generate an avatar capable of communicating using the knowledge that person had, Tolosana said. The world aspires to that level of technology, which is moving slowly in Latin America.
“AI will begin to impact from before we are born until after we die. It is an evolutionary issue at a Darwinian level,” he summarized.
According to the executive, the efficiency of AI is evident in its ability to provide solutions to problems with sufficient quality and equivalent to the performance of interns, as long as it has the necessary information. This is the reason for companies' investment in developing their own language model (training AI), which costs them at least US$100 million annually in electricity consumption, he noted.
In Latin America, according to the most recent Microsoft study ('AI Acceleration Factory') , companies in Argentina, Colombia, Chile and Peru plan to increase their investment in AI every two years, and 75% believe they will achieve the return on that investment in 12 months, with US$2.8 for every dollar invested. However, their initiatives do not have the same momentum as in other parts of the world due to the lack of talent and data governance [in form and structure], added Giovanna Cortez, general manager of Microsoft in Peru . 55% of companies use generative AI at basic levels, the same study found.
The race to apply AI in services offered is crucial to survive and stay ahead of people's needs. In this context, Tolosana highlighted the importance of tools, such as Azure Cognitive Search, a cloud-based search service that allows a person to integrate advanced search capabilities into their applications. In short, this service offers full-text indexing and search capabilities, as well as the ability to enrich responses through the use of AI.
"By using Azure Cognitive Search, organizations can take advantage of the scalability, security and reliability of the Microsoft cloud. Additionally, data stored at the blockchain is processed by AI (Azure OpenAI - GPT), allowing for greater automation of processes and improves the accuracy of results," he said.
The less information is transferred between different points, the lower the risk of exposure and manipulation. Cybersecurity, Tolosana continued, is a necessity today, especially considering that in April researchers at the University of Illinois Urbana-Champaign discovered that OpenAi's GPT-4 is capable of taking advantage of 87% of a list of vulnerabilities described in the list provided by the NIST (acronym in English of the National Institute of Standards and Technology in the United States).
With the wave of data voluntarily provided to these AIs, it is necessary for companies to know where their data is and “group” it, since it is more of a leadership problem than a technological problem.
“Before applying AI, companies must be clear about their roadmap [where to execute it], and be aware that the training to be provided must be at scale. All employees are involved,” Tolosana concluded.
