Just under two years ago, when ChatGPT launched, most people viewed artificial intelligence as a technological curiosity. Today, we are witnessing a veritable digital arms race, with the world\u2019s greatest powers vying for control of this transformative technology.<\/p>\n
One of the hottest topics of the last 24 hours perfectly illustrates this scenario: Big US tech companies are using the \u201cChinese ghost\u201d to try to loosen copyright laws in the US<\/strong>. According to UOL report<\/a>, companies like OpenAI, Google, Microsoft and Meta are pressuring the US government to allow freer use of copyrighted content in training AI models.<\/p>\n The central argument? \u201cIf we don\u2019t do this, China will overtake us.\u201d It\u2019s the old national security rhetoric applied to technology. It\u2019s no wonder that China has a national plan to achieve AI sovereignty by 2030, while the US is still racing to formulate its \u201cAI Action Plan.\u201d<\/p>\n This battle goes far beyond technology itself \u2013 it\u2019s a battle for economic, military and cultural power. The country that dominates AI will have enormous competitive advantages in virtually every field. And the tech giants know this.<\/p>\n Meanwhile, the race for ever more powerful models continues in full swing. Last weekend, Meta launched its new Llama 4 model, which includes two variants: Scout and Maverick. According to report by Poder360<\/a>, both versions are multimodal \u2013 that is, they can process and integrate different types of data, such as text, video, images and audio.<\/p>\n What is noteworthy is that Meta maintains its open-source strategy, unlike OpenAI and other players in the market. This approach can accelerate the democratization of AI, allowing more developers and companies to create applications based on advanced models.<\/p>\n Additionally, Meta also introduced the Llama 4 Behemoth, described as a \u201cteacher AI\u201d that teaches other smaller models. As TeleS\u00edntese article<\/a>, this approach results in significant gains in reasoning tests.<\/p>\n However, there are controversies: according to the Esmael's Blog<\/a>, Meta would be manipulating benchmarks, since the version of Maverick tested would not be the same as the one made available to the public. This type of situation raises important ethical questions about transparency in the AI industry.<\/p>\n \u201cWithin a decade, AI will do what humans do.\u201d This statement is not mine, nor that of any futurist, but that of Bill Gates, co-founder of Microsoft. According to Weather report<\/a>, Gates believes that AI will completely redefine the role of humans in areas such as work, education and medicine.<\/p>\n On the one hand, he sees opportunities, such as eliminating the shortage of doctors in underserved areas. On the other, Mustafa Suleyman, Microsoft\u2019s director of AI, warns that the technology could disrupt the workforce and replace jobs.<\/p>\n LinkedIn data reinforces this trend. A survey cited by State<\/a> suggests that 70% of the skills used in most jobs will change by 2030, with AI being a major driver of these changes.<\/p>\n For years, I\u2019ve been talking about the concept of CACACA \u2013 the six essential skills for the future of work: Creativity, Autonomy, Collaboration, Adaptability, Connection and Affection. In a world where AI takes over repetitive and analytical tasks, these typically human characteristics become even more valuable. In my mentoring with companies and professionals, I\u2019ve emphasized the importance of developing these skills to successfully navigate the new economy.<\/p>\n Not everything is rosy in the world of AI. As technology advances, criminals are also finding more sophisticated ways to use it for malicious purposes. Brazilian Post Office<\/a> warned about how AI is making cyber scams more convincing and dangerous, with the creation of fraudulent emails, cloned voice messages and videos manipulated with deepfakes.<\/p>\n Perhaps the most disturbing example is that of deepnudes, nude images produced by AI without the consent of the people portrayed. According to G1 report<\/a>, the emotional impact on victims is devastating, with reports of young people who no longer wanted to return to school after having their images falsified and distributed.<\/p>\n These cases highlight the urgent need for ethical and legal regulation of AI, as well as digital education to prevent and combat such abuses. We cannot allow technological advancement to translate into social regression.<\/p>\n Amidst so much worrying news, there are also positive examples of how AI can be used to improve public services. The Gramado City Council (RS) became the first in Brazil to implement an artificial intelligence called LegIA in its legislative process, according to Terra report<\/a>.<\/p>\n The goal is to increase efficiency, accessibility and transparency in the Chamber, allowing citizens to better understand the laws proposed by council members. It is a concrete example of how technology can bring public authorities closer to the population, democratizing access to information.<\/p>\n Initiatives like this show that, despite the challenges, Brazil has the potential to use AI in innovative and beneficial ways for society. In my lectures and consulting sessions, I often highlight cases like this to inspire public and private managers to see the transformative possibilities of technology.<\/p>\nThe race for more powerful models: Meta launches Llama 4<\/h2>\n
Bill Gates and the Future of Work: Ready for the Revolution?<\/h2>\n
The Dark Side of AI: Scams and Deepfakes Threaten Digital Security<\/h2>\n
AI in public institutions: the pioneering case of Gramado<\/h2>\n
AI in education: opportunities and risks<\/h2>\n