ChatGPT will top the mid-range first, and soon – the trending magazines on PC

There have been few major disruptive moments in the tech world since the advent of Google, social networking, and the iPhone. There are few digital innovations that can change the course of things so radically. That has changed. According to some observers, the launch of ChatGPT chatbot marks year zero for a new era in the digital world, and even in the business world. The artificial intelligence developed by OpenAI has impressed the entire planet since it appeared on the Internet at the end of November.
…

There have been few major disruptive moments in the tech world since the advent of Google, social networking, and the iPhone. There are few digital innovations that can change the course of things so radically. That has changed. According to some observers, the launch of ChatGPT chatbot marks year zero for a new era in the digital world, and even in the business world. The artificial intelligence developed by OpenAI has impressed the entire planet since it appeared on the Internet at the end of November. ChatGPT is very easy to use. You can ask any question you want and even ask him to write specific texts. Speeches, Poems, Recipes, Historical Facts, Articles: ChatGPT creates content on almost any topic or less and responds in a generally smooth, complete and “intelligent” manner. Almost like a human. Of course, a chatbot can basically give accurate answers to simple, realistic questions. But ChatGPT can also write a poem of the desired length and with certain words, or according to a specific style. It can also compose stories with superimposed characters or — we tested — write a press release with just a few bits of information. The results were surprising. Of course they are not perfect yet. Sometimes errors are made in statements or facts. Still, overall, there’s a largely logical and coherent structure, and even some form of reasoning, to the story. “This is the most exciting conversation between AI and humans that we’ve seen to date,” says Cyril Vaart, executive vice president of specialist consultancy EY Fabernovel. However, he is still a bit skeptical about the impact ChatGPT can have. Many have studied the application. Multiple tests have been done to see how ChatGPT can answer questions, hold a conversation, and generate more or less complete, technical scripts. And the world went crazy. Education, for example, where many pupils and students have entrusted all their homework to a chatbot. But most surprising and exciting are the results achieved by the chatbot in high-level tests. For example, international researchers from Chicago-Kent School of Law, Bucerius College of Law in Hamburg and the Stanford Center for Legal Informatics tested ChatGPT for exam questions to be admitted to the American Bar. According to them, the bot scored just over 50 percent. This is better than many students, while the chatbot has no training. Researchers are convinced that this type of artificial intelligence will soon be able to pass the prestigious test. Obviously, law isn’t the only subject in which ChatGPT can excel: He’s reportedly taken US medical exams, management exams, and even earned an MBA from the Wharton School of the University of Pennsylvania. To achieve these results, OpenAi has trained its GPT algorithm — short for Generative Generative Transformer — on a staggering number of texts across the Internet. The company does not disclose those sources. Observers refer to online databases such as Wikipedia, social networks, digital books, and anything that can be indexed on the net. According to specialists, the model is based on more than 300 thousand billion words, with a time horizon that does not exceed 2021. At least in the first version. Of course, the robot has no sense of what it is writing and does not understand the language. It is a purely mathematical model of the language. The machine only considers the context of words, neighboring words, and the probability of word-by-word occurrence. Humans are also involved in the process, for example to assess whether an algorithm is based on dangerous or sensitive content. For Laurent Alexandre, author of several books on artificial intelligence, the progress of language algorithms is far from trivial. Nor is it an easy development. He asserts that “ChatGPT is the biggest revolution since the advent of the consumer web in 1993”. “The applications in language recognition, understanding, and generation are revolutionary because a very large part of human, economic, and creative activity is the production of language content.” ChatGPT is a text and content generator. This is a job that many people do in many industries: from a project note to a PowerPoint presentation or press release to a Facebook promotional post. Currently, customers who call a bank, telecom provider or energy provider are greeted by artificial intelligence. “It works through a conversation tree, which is a different technology,” says Cyril Wart of Fabernovel. “If a machine doesn’t know how to answer, a human will flash a light.” In the world of copywriting and content creation, ChatGPT can turn the tables even more radically. For example, a marketing agency focused on social media requires its teams to prepare promotional briefs, plans and scripts on a daily basis. A fairly advanced chatbot, like that one from OpenAI, can create Facebook posts on demand. “ChatGPT will be used in copywriting and even strategically,” predicts Sabrina Pultow, who has long run social agency BeConnect. “This technology is going to do a lot of the work. The big advantage is that ChatGPT provides a basic structure to which you, if you know the work, can add a layer of human intelligence. But it will also be able to create content that can be used almost entirely, based on a good briefing.” , which saves a lot of time and is therefore effective. In a recent survey by Start-up Sortlist, 22% of employers in the financial sector would consider downsizing if they were offered ChatGPT. 51 percent of them look primarily at the marketing department. This reasoning can be extended to other sectors. “Few companies will continue to ask someone who costs $5,000 a month to do what ChatGPT can do in 17 seconds,” predicts Laurent Alexandre. “The report that is usually entrusted to a middle manager will be taken away from him. It will eventually be a problem for middle managers in the service sector. Middle managers will have to prove that they have added value, because they are very expensive.” According to the author, many sectors will be severely affected, And the legal staff will be at the fore. “We are heading for a major reorganization of businesses, education and even hospitals,” he warns. Alexei Greenbaum, director of research at the French Institute for Research on the Fundamental Laws of the Universe, was more moderate in a televised speech, without denying the impact of artificial intelligence: “Entire fields of work will change, but they will not disappear. There will always be doctors, judges, lawyers, journalists, etc., but the content of their work will change. This is also the vision of Michel Levy-Provençal, a French teacher specializing in innovation issues and founder of Brightness, a company that uses artificial intelligence. He acknowledges that “artificial intelligence can speed up content production.” “But it’s still basically a tool. AI itself doesn’t generate anything, and people have to demand it. The way tasks are given to AI will also be a skill. We’re actually increasing the value a little bit and gaining productivity.” And even if the results of ChatGPT Still far behind human production, “at the rate this is progressing, it’s reasonable to think that it will be difficult to distinguish machine from human three or four years from now.”