Generative Artificial Intelligence created by Open AI has already received a contribution of US$ 10 billion from Microsoft and can help reduce costs for agencies
This text was not written using ChatGPT technology. But it could! Created by Open AI, the tool is ready to revolutionize the world of communication and, according to experts, go beyond searches using the SEO method, Wikipedia, etc. It promptly delivers the suggested demands to this Generative Artificial Intelligence that promises to be the biggest step ever taken since the industrial revolution in the last century.
This robot can think, so to speak, like never seen before in this technological ecosystem, with real-time answers to any type of question. And for free! Can be used for Q&A, text summarization, translation, sorting, code generation, etc. Some people believe that it can replace Google. To be?
In less than a month after its launch, at the end of 2022, it has already received an investment of US$ 10 billion from Microsoft. And also funds. There are already signs that ChatGPT will enter phase 4 in the second half of this year.
In the words of Alex Winetzki, CEO of Woopi and director of R&D at Stefanini Group, ChatGPT aims to imitate human intelligence, but it has something that robots cannot contemplate: consciousness. According to him, the system does not know what it is doing even if it reproduces human characteristics efficiently.
Generative AI already shows how it will change the world. It advances into the territory that we always thought was beyond the frontier of machines: human creativity. Dall-e, Stable Difusion, ChatGPT and many other technologies that came to light in the last months of 2022 show that robots now know how to create as well or better than most humans, ponders the Woopi executive, who continues.
ChatGPT just the most powerful language filler ever. He has no commitment to reality, for two reasons. The first is that he doesn’t understand reality. That is, he has no personal history, no sense of connection to the physical world, self-criticism, moral or behavioral restraints, personal relationships, ethics or common sense. What it does is simply look for materials already produced by someone, somewhere, and cleverly repackage them, based on parameters given by the user, explains Winetzki.
In fact, the initial logic of Generative AI was to look for the next word that would make the most sense in a sentence, that’s all. An assembly game that grew with computational power for billions of parameters (a parameter is a small block of text digested on the internet), leaving the result increasingly sophisticated. The challenge, therefore, is to be able to create efficient models of Generative AI with less training parameters, which allow more flexibility in the use of the tool. That said, and these are technical notes, the commercial and social disruption potential of what we’ve seen so far is profound, Winetzki said, emphasizing that ChatGPT is a Deep Learning model, but that it is still a black box.
ChatGPT generation 1 parameters amounted to 117 million; in 2, it jumped to 1.2 billion; in 3, the expectation of 175 billion; and in the 4th, the forecast by Andrew Feldman, from Cerebras and partner of Open AI, in an interview with Wired, that there will be 100 trillion parameters. The fact is that ChatGPT offers multiple functionalities for the business world: small, medium and large, especially retail (physical and digital).
And how can advertising benefit, or not, from this Generative AI resource? How should agencies understand Chat-GPT? Who answers Daniel Martins, head of operations and transformation at Galeria.ag. Absolutely in all disciplines that we have today it is possible to find some benefit. There are already dozens of tools that have excellent potential and adoption curve and can add activities ranging from adaptation and unfolding of image, text, audiovisual content, to some that already offer some benefit for activities in more abstract stages, diagnoses Martins.
He recalls that there are risks to publicity: In addition to the issue of privacy of information, which is always a critical issue when we work with the use of data that is fundamental for any AI to work , we also have important issues related to unconscious biases of these platforms and even potential intellectual property risks. I think the biggest challenge will still be how we will form a generation of talent that knows how to connect the best of technology with human potential and creativity. Orchestrating all of this will bring impacts to our business on a scale we haven’t experienced yet, advances Martins.
Read the full story in this week’s issue.
(Credit: Zac Wolff on Unsplash)
