What is ChatGPT and why do some see it as a threat?

[ad_1]

System created by the company OpenAI, co-founded by Elon Musk, was considered a new leap for artificial intelligence and is seen as a disruptor in several areas, such as education and creativity. Meet ChatGPT, the technology that went viral for having an answer to (almost) everything Last November 30th, while the world was starting to get into the end of year mood and had its eyes more focused on the Qatar Cup, a program was launched that it was soon considered a step forward in the progress of artificial intelligence: ChatGPT. But the consequences for humanity of its large-scale use have yet to be fully understood. The evolutionary leap of this new system is in the power to generate content in a very coherent way and also for being able to “sound more human”. Despite the errors and shortcomings that still exist, the capacity already demonstrated by the program — and its potential to develop even further in the long term — has provoked not only admiration, but also some fears. Just look at the amount of reviews that associate ChatGPT with the word “threat”. Many feel that the program looks too convincing when mimicking the speech of a human being and foresee problems. These are questions about the possibility of a strong rupture in areas such as human creativity, learning and education, work, digital security and democracy itself, as stated in a recent article in The New York Times. According to the authors, what used to be a person expressing his political opinion may now be just a robot that artificially generates an argument. What is the program? ChatGPT is, basically, a virtual robot (chatbot) that answers various questions, performs tasks in writing, converses fluently and even gives advice on personal problems (although there is a caveat that it does not have this objective and that past guidelines are generic. But he almost always responds to the user for advice). The possibilities for generating content are immense. He can, for example, teach you how to prepare a stroganoff — and with a different twist on the recipe, if you specify —, give tips to get a job, write poetry, academic works or a model of judicial power of attorney and also compose a letter Reconciliation letter for a friend you’ve grown apart from. ChatGPT responds in a matter of seconds to a random request such as “write a poem about artificial intelligence in the style of Carlos Drummond de Andrade” — the quality of the result, depending on the critic, can be quite debatable. According to SEO.ai tests, ChatGPT is available in almost 100 languages ​​(including Brazilian Portuguese), but the performance of the model varies according to the language (it works better in English). The system was developed by OpenAI, a company founded in 2015 in the USA by Sam Altman (today its main figure) and by the ubiquitous Elon Musk (who left it in 2018 because he considered that there was a conflict of interest with his main enterprise, the automotive company Tesla). Five days after its launch, ChatGPT reached over 1 million users. These interactions are being used to train and develop the model. OpenAI says it will be free and open to everyone (although, ironically, you’ll need to prove you’re human to log in) during this “testing and researching” stage — leading experts to speculate on future types of monetization for the tool. The company also warns that, during this period, the software “may occasionally generate incorrect or misleading information” and that its data history is limited to 2021. Despite having been identified as a possible threat to Google’s hegemony as a facilitator of information on the internet , the system still makes serious mistakes, like saying that Brazil has already won at least five Oscars (in fact, the country has never won a statuette). Why is ChatGPT considered a breakthrough for AI? Powerful text-based artificial intelligence programs work by storing gigantic amounts of data (with an emphasis on words and conversations in this case) and with algorithms to predict the best chaining of a sentence. These are called large language models (LLM). Unifesp professor Álvaro Machado Dias, a futurist and neuroscientist, explains that when training software, “questions are asked to the chatbot such as ‘what is a cylinder?’ and technicians elaborate their own answers. If the chatbot’s answer is not correct, the correct ones are inserted in the system to teach it. This is passed on to other situations automatically”. Despite already using a modality that manages to understand the context of the use of words, allowing better concatenated texts, previous programs did not respond so well to the user or still sounded very artificial. ChatGPT learned to talk in a way closer to a human. Machado Dias says that the differential of this program is the use of a technique that understands how language works: the reinforcement of learning through human feedback (RLHF). Engineers apply “reward” and “punishment” methods to teach the system the most desirable forms of interaction. It’s a fine-tuning process. “In practice, the engineers order the responses given by the algorithm according to their relevance and encourage the program to learn the preferences listed in the ordering to increase the relevance of textual productions. The result is given in the form of texts that seem deeper and more significant than those of the alternatives”, says the Unifesp professor. ChatGPT has also been trained to admit mistakes, challenge incorrect assumptions and reject inappropriate requests. But a professor at the University of California managed to get the system to write programming code to say that only white or Asian men make good scientists. OpenAI claims that although it has made efforts in this regard, the program “will sometimes respond to problematic instructions or exhibit biased behavior”. According to the company, the data collected in this stage of testing will serve to improve the system. Is it a threat to learning and creativity? The threat of disruption already hangs over work and employment. Fields that depend on text, like journalism, could be greatly modified – and vacancies could disappear forever. ChatGPT’s competence in generating codes is also already raising questions in a relatively new sector, programming. But one of the areas that has been realizing the potential problems of ChatGPT is precisely one of the most affected by the arrival of new technologies: education. The temptation among students to use the program to find ready-made answers to their tasks led New York to make a quick decision: just a month after its debut, the system was banned from schools and public devices in the American city. OpenAI has been working on a kind of digital watermark that identifies that the content originated in ChatGPT. There are already algorithms that calculate with good accuracy the probability that a text was made by a chatbot. In addition to “copy and paste”, there is the fear of structural impacts on human learning. For example, will the cognitive exercise of writing an essay with a beginning, middle and end, concatenating ideas in a cohesive way, be affected? “I’m very concerned with the algorithmization of thought, which is the change in our understanding and relationship with the world as a result of interaction with AI”, says Machado Dias, from Unifesp. “I believe this will be the biggest change in mindset in all of modern history. It is worth noting that the human brain has been slowly reducing in size, as a result of technological development, for over a thousand years. This process must accelerate. Thus, we will become increasingly sophisticated from a technical-cultural point of view, but also more limited from a neurocognitive point of view.” Martha Gabriel, author of the book Artificial Intelligence: from Zero to the Metaverse and professor at PUC-RS and SP, states that it will be necessary to adapt to new times: “When a technology starts to do something better than a human being, it’s no use wanting to compete with her in it. From that moment on, the human skills that become valuable are: knowing how to use that technology to its maximum potential and doing what the technology cannot do”. “What makes the difference in this context are no longer the answers, but the questions. You have to know how to ask. To know how to ask, you have to know how to think critically”, she says. For Yuri Lima, a researcher at the Laboratório do Futuro at the Federal University of Rio de Janeiro (UFRJ), “teaching itself needs to be adapted to stimulate learning that recognizes students as cyborgs increasingly integrated with new technologies”. “This demands that teachers also know how to use these same technologies and integrate them into their activities. From the moment that projects, activities and homework become more complex and integrated into the current universe in which students exist, the motivation is no longer affected by this technological issue.” Another issue revolves around the future of human creativity and content production that is not based on artificial intelligence. Ten days after the arrival of the OpenAI system, a designer in San Francisco (USA) managed to create in just one weekend a children’s book with text and illustrations made with ChatGPT and MidJourney, a program that produces images by providing descriptions. “An important point still to be addressed is the issue of plagiarism. Since the training of models such as ChatGPT is based on texts available on the internet such as news, books and blogs, their answers may bring ideas published by certain people without them receiving credit for it”, says Lima. “In creative areas, this recognition of the authors is considered important. Furthermore, not knowing the sources makes it difficult to recognize biases or even lies, as in fake news.” Machado Dias points out that “creativity emerges from combinations that are simultaneously unusual and relevant. As algorithms are devices for generating combinations, it is expected that the creative impulse will be reduced”. “On the other hand, as algorithms perform their tasks, they tend to generate patterns that we didn’t even imagine as possible, expanding our combinatorial understanding, that is, our creativity.” Martha Gabriel goes in a similar vein: “Technology can be an incredible tool to amplify our thinking, as we can in a short time test countless hypotheses, formats, solutions to refine our hypotheses and improve our questions”. “However, this can also be a very big threat for those who use these systems blindly, without criticism or questioning of morals and ethics. This is not just an individual risk, but for all of humanity.” – This text was published in

[ad_2]

Source link