Brazilian Tech Firm Wide Labs Unveils Amazônia IA at National Conference
Porto Alegre, Brazil – Wide Labs, a rapidly growing tech company based in Porto Alegre, has announced the development of Amazônia IA, one of the first language models (LLM) created specifically for Brazilian Portuguese. This groundbreaking initiative will be showcased at the National Conference on Science, Technology, and Innovation in Brasília, where President Lula is set to unveil a national plan to boost artificial intelligence (AI) development.
Amazônia IA: A New Chapter in Brazilian AI
Named after the Amazon rainforest, Amazônia IA is described as robust, delivering high-precision results. Although technical specifics such as the number of parameters involved remain undisclosed, the technology promises significant advancements in Brazilian AI capabilities.
Nelson Leoni, CEO of Wide Labs, stated, “É ciência brasileira na veia” – emphasizing the purely Brazilian nature of the project. Leoni, an ex-Army officer and former marketing strategist at BTG, was injured during a UN peacekeeping mission in Haiti and later transitioned to the tech world.
Wide Labs collaborated with industry giants Nvidia and Oracle for this initiative. The Amazônia model will operate through two Oracle data centers located within Brazil, ensuring compliance with local data protection laws, specifically the Lei Geral de Proteção de Dados (LGPD).
Competitive Edge and Future Plans
Wide Labs aims to differentiate itself from major US tech companies by offering services in Brazilian reais, promising substantially lower prices. Following the example of Big Techs, Amazônia IA will feature a free conversational bot and subsequent premium services. “Vamos cobrar em reais, e não em dólares,” Leoni pointed out, highlighting price competitiveness.
Moreover, Wide Labs plans to launch smaller, specialized models tailored to sectors such as legal, communications, and medical fields. This diversification of their AI offerings aims to capitalize on sector-specific needs, ensuring broader application and adoption.
Pioneering AI Solutions in Brazil
The company’s innovation credentials are underscored by their previous project, bAIgrapher, an AI biographer designed to aid Alzheimer’s patients in cognitive function preservation. This project, still in experimental stages, has already garnered international accolades, including a Bronze Lion at Cannes.
An Emerging AI Ecosystem
Wide Labs isn’t alone in the Brazilian AI landscape. Maritaca AI, another pioneering firm led by Rodrigo Nogueira from the University of Campinas (Unicamp), has developed MariTalk, a native ChatGPT-like model. Initially reliant on pre-trained external models, Maritaca AI now prides itself on its proprietary system, Sabiá, trained entirely within Brazil.
Nogueira emphasizes the model’s localization benefits, enhancing its utility for Brazilian users by aiding in tasks such as consumer complaints to Procon, tax filing, and investment evaluations. He foresees an AI battleground where models compete based on usability and relevance to local contexts, an area he feels lacks consumer protection tools.
National Strategy for AI Development
Brazil’s AI ambitions face challenges, notably a shortage of qualified professionals and a need for enhanced computational power. Addressing these obstacles, President Lula’s AI policy includes the construction of a supercomputer designed to support Brazilian researchers, potentially revolutionizing the nation’s capabilities in this technological frontier.
As Brazil edges further into the AI domain, firms like Wide Labs and Maritaca AI exemplify the local innovation spirit, driving forward with models uniquely tuned to Brazilian needs and conditions. The launch of Amazônia IA marks a significant milestone, heralding a new era of technological autonomy for Brazil.
For more information about Wide Labs, visit their official website.
