[ad_1]
Chat GPT it may not tell you much but if you surf the web every day you will surely have come across this name and wondered what it is. First of all, know that ChatGPT is the name of the organization dedicated to the research and development of artificial intelligence Open AI he gave to his new chatbot when he officially presented it to the world at the end of November this year.
Chatbots are nothing more than those software capable of simulating human conversations responding to sentences written or spoken by users. They have been used for years now because they allow people to interact with digital devices and websites: the most striking examples are undoubtedly those that appear on e-commerce sites to help customers who need help. In fact for no chatbot put on the market so far comes close to the capabilities that ChatGPT is demonstrating. A very important premise before understanding how it works and what it can really do.
ChatGPT: what is it?
Let’s start by knowing what ChatGPT is. As mentioned at the beginning, nothing more than a real prototype of Artificial Intelligence developed by OpenAI, the same that also launched Dall-E, and which is able to understand human language and entertaineven very complex conversations. For the record, OpenAI was founded in 2015 by Elon Musk and other Silicon Valley investors. The idea was to to advance digital intelligence so that it could benefit mankind. To date, the CEO of Twitter, Elon Musk, is no longer on the OpenAI board and has even distanced himself from everything OpenAI is doing and where he wants to go.
ChatGPT: How does it work?
ChatGPT completely free although in this case to use it necessary register by creating a profile directly on the OpenAI website on this page. On a technical level the entire conversation system is based on the technology that OpenAI calls GPT-3.5 and none other than an artificial intelligence that has been ”trained” with a myriad of texts such as newspaper articles, books, documents and more, to allow at any time to be able to converse freely in an articulated way being able to answer as if one were in front of a person.
The idea of OpenAI is to make ChatGPT as ”conversational” as possible precisely allowing everyone to be able to use it and above all to have the sensation of being able to converse with a person and not so much with an Artificial Intelligence such as instead. The naturalness with which ChatGPT responds makes it different from the chatbots currently on the market. In fact, in the answers that we can obtain there is a syntax and a construction of the sentences that are practically indistinguishable from the human one above all because ChatGPT allows you to answer in a precise and reliable way with respect to the context of the question but above all it allows you to have a ”remember” of what has previously written with the user.
Moreover ChatGPT very sincere with the user because when he cannot answer a question he openly declares it. Not only because he corrects himself in case of errors and above all he has the ability to avoid misunderstandings and also to avoid explaining to people how to do, for example, dangerous or illegal things such as stealing a moped or creating a bomb. Although in this case some examples online have revealed that the chatbot can still answer if the question is somehow ”bypassed” by the user.
In fact ChatGPT allowed some programming code to be written and correctedhas wrote songs in versehas composed pieces for pianohas explained how to cook Carbonara and still allowed to correctly diagnose diseases starting from a series of precise symptoms or enjoyed making jokes or inventing fairy tales with the characters you want. On the net we can really find everything and the level with which he is able to respond to conversations is incredible.
ChatGPT: what could be its future?
ChatGPT’s ability to converse is clearly fascinating and at the same time a little terrifying. Many see the new OpenAI chatbot capable of replacing Google in the coming years to do searches even if at the time of writing you should know that OpenAI has updated its artificial intelligence to 2021 and therefore if we ask for current events we will not get effective answers. In addition to this, however, many bet that such a system could easily replace human jobs such as those of a classic call center or after-sales service.
Not only because even those who carry out artistic activities could be supplanted in the future by ChatGPT. The most striking example is being able to write technically correct articles starting from little or a lot of information. Clearly still too early to throw zero on the fact that ChatGPT could steal the work of humans above all because OpenAI itself admits that artificial intelligence is often unable to achieve what is requested or still makes too many mistakes. clear for if artificial intelligences will continue to grow and improve, the problem will have to be addressed as soon as possible.
Its danger? On the net there are examples of how ChatGPT has been able to write extremely convincing essays starting from simple discussion data. There are also examples of articles on specific and even highly technical topics that have been drafted in a precise way with a result that has surprised even the most skeptics. And precisely this ability to write texts in an easy but equally effective way could be the cause of one in the future radical spread of disinformation such as to be able to manipulate even people on a large scale without realizing that they are not real human beings but artificial intelligence. We know well how in general platforms have always struggled to report which users are actually human beings and which instead of bots. In this case due to the obvious difficulty of cognition and if bots or artificial intelligences really took over the damage could be potentially high.
The idea of a secure, harmless but also useful ChatGPT it might come trying to figure out its blind spots that might lead to misuse for malicious purposes. Certainly the OpenAI engineers are working to fill these gaps and it is also probable that in future versions ChatGPT may no longer be free. Yes, because every time users ask something to the artificial intelligence, Sam Altam, founder of OpenAI, said that it costs the company a few cents, so the amount of questions ChatGPT is asking is basically costing hundreds of thousands of dollars considering that in the first 5 days alone it was tried by over a million people.
.
[ad_2]
Source link
