ChatGPT artificial intelligence is panicking the world of new technologies

Some say Google’s days are numbered. ChatGPT, the first conversation interface with an artificial intelligence, is available for free, and the public/machine dialogues offered for less than a month are disconcerting. Do you have a question? ChatGPT will answer you very clearly.

ChatGPT is the addition of a tool for cat (conversation, in English) and a Generative Pre-trained Transformer (GPT), a generative artificial intelligence of text. You are therefore offered to converse with an artificial intelligence trained to answer any type of question, and it works remarkably well in French.

In its previous versions, the smart tool first absorbed a huge amount of available texts, via machine learning. Through an operation of fine tuning, with humans this time around, the machine later improved to accommodate the novelty and difficulty of the conversation game with human brains. (Yann Le Cun from the College de France explains here the modalities of learning a machine)

Origin of ChatGPT, OpenAI. This company was founded in 2015 by Sam Altman (he just San Francisco and exactly the impressive start-up incubator Y combinator ) and other experts, including the no less famous peter thiel and Elon Musk. The latter withdrew from the project in 2018, because according to him: ” OpenAI was born as an open-source, non-profit organization and neither of these two adjectives is true today. The company received $ 1 billion from Microsoft in 2019, open source has closed and we bet that a lot of money is at stake around this new technological use. Open AI had launched a few months ago DALL E an artificial intelligence which, without conversation, could create images from text. The results and the creative ability of the machine have been a tsunami among graphic designers and image makers.

More than a million users in five days

Sam Altman, OpenAI’s chief executive, says ChatGPT hit the one million user mark in five days. Facebook had taken ten months to obtain such a membership rate. And enthusiasm is overflowing for this new human/machine conversation, Silicon Valley is in turmoil, as are tech fans and analysts…

ChatGPT is an interface for the moment very simple to use. After an appropriate hello, the machine asks us what we are wondering about. Then emerges the curious psychological need to dialogue with a machine, not necessarily obvious at first sight, but very quickly the conversation turns out to be fluid and rich. You just have to try to formulate questions to get answers.

Didactic, educational, ChatGPT answers all scientific questions, history questions, etc. We understand very quickly what this artificial intelligence explains to us, on any subject. ” His answer to the question: “what to do if someone has a heart attack” was incredibly clear and relevant. “, explained to AFP Claude de Loupy, director of Syllabs, a French company specializing in the automatic generation of text. And when a user mixes ChatGPT’s answers with the illustrations provided by Dall E, it results in a children’s story in just a few questions.

West France also tried his hand at the exercise, asking ChatGPT to detail the recipe for making pancakes. The answer is unstoppable. We also find in Numerama that the machine is able to write a text on any subject according to the style of a known author. The results are always very fast and stunning for lovers of Marguerite Duras like those of Eminem.

The Reddit thread on some conversations with the machine is uplifting and funny.

Many struggled to find errors in some answers. There are indeed some, but ChatGPT is a machine that learns from its mistakes. So the more of them there will be to trap the machine, the more it will learn somehow… For computer scientists, the answers offered by ChatGPT seem dizzying, because writing code, verifying it or correcting it, seems extremely easy for this machine. Just like detecting security vulnerabilities. It’s abysmal and the repercussions of this very recent dialogue with the machine are only just beginning. How to say, it’s going too fast?

Dialog around computer code with ChatGPT. © Open AI

Such impressive technology that makes Paul Bucheit, creator of Gmail, say: “ Google is finished. He is not the only one to think so, because, when the search engine only “indexes”, ChatGPT offers reflection and solutions adapted to the request. ” What the search engine did to the Yellow Pages by making them obsolete, ChatGPT will do to Google adds Paul Bucheit. And he demonstrates it in a series of tweets.

“This technological revolution is irreversible”

Artificial intelligence is the possibility of changing many things in everyday life proclaims Sam Altman of OpenAI. ” This technological revolution is irreversible “, he writes in a blog post and he insists on the incredible change that this tool will cause in our lives. He is also cautious on ChatGPT which is in fact only a ” early demo of what is possible “. Soon, “ you will be able to have helpful assistants who talk to you, answer your questions and give advice. Later you will have something that will do the tasks for you. »

In use, we really feel like we are talking to a very, very intelligent friend, almost funny, and always asking our questions… A bit like having a passionate conversation with his GPS navigator or with the robot Hal in the film 2001 a space odyssey, and always with the same feeling of vertigo: the machine never rests, thinks faster than us and will it overtake us? “ A few years ago, chatbots had the vocabulary of a dictionary and the memory of a goldfish. Today, they are much better at reacting consistently based on request and response history. They are no longer goldfish notes Sean McGregor, a researcher who compiles AI-related incidents on a database.

Has the ChatGPT wolf entered the technological fold? ?

Doesn’t this type of intelligence represent a certain danger? Paul Kedroskyexpert, fears him, he uses the analogy of robotization to express his distrust, even his fear. Thus, robotizing a factory, even if it causes a shock in working methods, does not risk transforming it in the long term. Mechanical robots are not made to seek to improve work themselves, or even to shake up habits in the factory and modify the company’s external purchases. An artificial intelligence like ChatGPT could offer such a revolution in work, and with generosity. He already expresses doubts in the work of his students who use this text generator for their written work. All this is extremely dangerous according to him: ChatGPT must be removed from the market immediately. »

Has the ChatGPT wolf entered the technological fold? Sam Altman, co-founder and boss of OpenAI, denies it: It’s interesting to see people wondering if AI (artificial intelligence) systems should behave the way users want them to or the creators intended them to.. But what has he planned? Or rather what does this artificial intelligence predict for our technological future? Can we trust him to make decisions in our lives? ” The debate on what values ​​to give to these systems is going to be one of the most important that a society can have. supports Sam Altman.

Some talk about a next paid version of ChatGPT? Speculations, based on a GPT4 version which would already be in beta test, are already making a lot of noise… So? To log in and test (it’s easy and it’s free) it’s here. The chat, the conversation tool, is here. It’s up to you to dialogue with the machine:

You : Good morning !

ChatGPT : Good morning ! How can I help you today?

We wish to give thanks to the author of this post for this awesome material

ChatGPT artificial intelligence is panicking the world of new technologies

Check out our social media accounts and also other pages related to them