The transformative force that the generative artificial intelligence (AI) It is having in various fields: from the creation of content and design to the research and development of documents, strategies and more. Its ability to produce texts, images, audio and video with similarity to human creation opens a range of unprecedented possibilities.
The promise of greater efficiency, amplified creativity and personalized solutions promotes an increasingly widespread adoption of these technologies. In Colombia, according to a Microsoft study, 82 percent of companies plan to mobilize part of their resources and investments in AI, both in technology and training and optimizations of your workforce.
Microsoft also revealed last year that about 80 percent of people use AI for tasks in their work, while a GAD3 study, published in November, showed 84 percent of students in the country use the generative the educational work daily.
However, the ease of use of generative AI can induce errors due to excess confidence or ignorance. Without a critical look, This powerful tool can become ineffective or risky.
Photo:stock
Common errors
One of the most frequent is Assume that the information generated by AI is always accurate and truthful. While these models have trained – and train daily – with vast amounts of data, they can also replicate biases of that data or generate incorrect information, known as ‘hallucinations’.
A study published in Nature magazine revealed that even the most advanced language models can produce incorrect factual information in a significant percentage of their answers. Therefore, It is essential to verify the information given by the generative AI with reliable sources before using it to make decisions or disseminate it.
A second common error lies in the lack of specificity in the indications given to an AI. The more ambiguous the Prompt (instruction), the less likely it will be that the AI generates a result that meets the user’s needs. The correct way is to indicate in detail what you want, how, where to get inputs (even delivering documents or links), specifying the format, tone and any other element.
The third mistake is to forget that you are ‘the human’ in the equation: Depending excessively on generative AI without applying critical thinking is a great mistake. While these tools can automate tasks and generate ideas quickly, they do not have the reasoning capacity, the understanding of the in -depth context or the ethical judgment of a human being. The user’s role remains essential to evaluate, refine and contextualize the production of AI.
Photo:stock
Finally there is the most important: forgetting to protect sensitive data when using generative the AI, especially in free versions. Enter personal or business information can make it public or expose it to third parties, Not to mention that the platform uses them to train, free, their AI models. Therefore it is crucial to review the privacy of the platforms and not share sensitive data without clear guarantees.
A general recommendation in this regard is the measured use of AI: each consultation made generates an important Energy and Water Expenditure (To cool the data centers that process these language models), so dedicating themselves to ‘playing’ is a mistake that even the platforms ask to avoid.
The most popular options
Galaxy AI y Apple Intelligence: present in Samsung Galaxy and iPhone and iPad cell phones. They offer functions of editing images and execution of daily tasks.
ChatGPT (chat.openai.com): It is the most commonly used conversational, you can create coherent text and answer questions.
Google Gemini (gemini.google.com): Another AI capable of solving doubts, creating texts, documents, and more.
Microsoft Copilot (Copilot.microsoft.com): an AI also integrated into Windows, Office, Outlook, for the generation of texts, etc.
From the-e (Openai.com/dall-e2): Chatgpt ia to create images.
Suno Ai (Suno.ai): It is the popular AI to create songs.
José Carlos García R.
Editor Multimedia
@Josecarlostecno
