Recently, ChatGPT has become extremely popular. This incredible artificial intelligence chatbot has become a real discovery in the tech world.
I don’t remember such a hype since the explosion of cryptocurrencies, and maybe even since the debut of the first iPhone. ChatGPT has been on the market for only 2 months and has already caused chaos and upheaval in a large part of the consumer technology market.
It’s hard to believe that the updated version of the OpenAI chatbot debuted only on November 30, 2022. Within 2 months, ChatGPT not only changed the world of online business but also almost destroyed the foundations of the world of education and entertainment.
The most important facts about the ChatGPT debut
ChatGPT is a language model trained using a machine learning technique called Reinforcement Learning from Human Feedback (RLHF) that can engage in dialogue, answer follow-up questions, recognize errors, question false premises, and reject inappropriate requests. ChatGPT has been trained on examples of interpersonal conversations, which helps to make the dialog more natural.
OpenAI’s new ChatGPT chatbot has taken the Internet by storm and shown how much fun it can be to talk to artificial intelligence. However, despite its impressive capabilities, it still makes a lot of mistakes, which may reassure those who fear that artificial intelligence will take over our jobs and then our lives.
In recent weeks, screenshots of conversations with ChatGPT, the latest artificial intelligence model developed by research firm OpenAI, have gone viral on social media. ChatGPT has been asked to tell a joke, write an episode for a TV series, an exam essay for an eighth-grader, or even customize computer code.
More than a million people have started talking to artificial intelligence, and while they haven’t always gotten satisfactory, meaningful answers, ChatGPT has proven to be a pretty good writer, as well as an interesting conversationalist who can answer a tough question and tell a joke. With the latest updates to other OpenAI software, DALL-E, which is used to create graphics, and Lens AI, a platform that creates digital portraits, everything points to AI being able to compete with human abilities. At least in some areas.
Just 3 weeks after ChatGPT debuted, Google turned on the “red crisis button” mode. The company’s current CEO Sundar Pichai asked for an urgent meeting with the creators of the Google search engine, Larry Page and Sergey Brin. Pichai fears that OpenAI’s artificial intelligence tool can pose a real threat to Google’s hegemony in the search market.
Especially since ChatGPT is a much more versatile tool than Internet search engines. For example, it can create posts on social networks, write scripts for TV shows, recipes for vegetarian dishes, music, or even sermons for priests.
The creators of ChataGPT have been cooperating with Microsoft for a long time. It was the company run by Satya Nadella that invested a billion dollars in OpenAI in 2019. In January of this year, Microsoft promised to give another $10 billion for the development of ChataGPT. It is assumed that artificial intelligence will come to the Bing browser in March. A powerful earthquake is brewing on the Internet market. Google Chrome’s position has been shaken and is under serious threat. All of this is turning into a struggle between corporations, worldviews, ecosystems, etc.
ChatGPT has gained incredible popularity among ordinary users. So has Midjourney, which is a breakthrough when it comes to the use of artificial intelligence by “ordinary” people. In the end, each of us sees the potential of AI-based programs and the great results that can be achieved with its help. Of course, we know that AI does not “create” something out of nothing; the program learns from huge databases of source materials. This, in turn, means that, as with Midjourney, using ChatGPT, for example, at school, is not only cheating, but also most likely unconscious plagiarism. Therefore, while we encourage you to try out chatGPT, it is better not to use the results of its work.
Let’s find out what exactly ChatGPT can do. Let’s try to figure out in which areas of activity it has already become quite advanced, cool, and successful.
Writing school essays
ChatGPT is disrupting the foundations of general education. It turns out that the AI tool is excellent not only for homework, but also for exams, essays, and even large graduation papers.
Such features of ChatGPT have proved to be extremely popular in schools and universities, where students are beginning to use it en masse. ChatGPT can combine knowledge from many sources and present it in a coherent way, while maintaining the sequence of causes and effects. This means that teachers are usually unable to guess whether a text was created by a student or a machine. The creators of the tool have promised to help institutions detect such fraud, but at the same time, no one can guarantee 100 percent efficiency of this method.
The problem is so serious that in the wake of the chatbot’s potential to help students, New York City decided to block access to ChataGPT on all devices and networks belonging to the city’s public schools. Since January 4, both students and teachers have been unable to connect to the chat.openai.com domain in schools, and the block covers not only school devices but also networks. Since then, the media has been reporting daily on similar blocking in many schools, not only in the United States.
However, given the fact that a simple cell phone is enough to use all the features of this bot, these bans are symbolic. Anyone who wants to will still use the bot. It is difficult, if not impossible, to block this possibility.
The creators of ChataGPT warn schools against panic blocking chatbots. The OpenAI CEO claims that such tools can be a useful aid in education. He even suggests using ChataGPT as a tutor that could explain specific issues that a student might have a problem with. In his opinion, these benefits outshine the problem of the student not creating written homework on their own, which ChatGPT can do instead.
One thing is for sure – there is a huge turmoil in the education system.
Creating content in foreign languages
There are situations when we need to translate a document or a test, or part of a text, into a foreign language. It’s good if you or one of your friends knows these languages perfectly, but this is not always the case. It is known that learning foreign languages requires a lot of effort, and not everyone wants to do it.
In this case, you need a qualified translator who will make the necessary translation for you, possibly for a fee. This situation can happen again and again, and sometimes it’s hard to solve it right there and then. But there is a way out.
After all, if we want to create a short note in several languages, we can safely use ChatGPT, which will not only create but also translate the message into several languages. It’s really fantastic. The whole process of completing a task in ChatGPT will take a little time. Of course, it is not without mistakes, but it is worth it.
Writing song lyrics
The creative process can sometimes be difficult, which I can confirm from my own experience. Sometimes you just don’t have the inspiration to write lyrics, and if you do, it’s sometimes hard to find the right words. This is where ChatGPT can help, as it can not only write a lyric on a specific topic, but also arrange a song in a certain musical style.
Recently, the Internet was stirred by the news that ChatGPT wrote a song in the style of Nick Cave. It was quite an interesting story. One of the fans of the rock band Nick Cave and the Bad Seeds used the ChatGPT chatbot to generate lyrics for several songs in the band’s style. He sent the finished lyrics to frontman Nick Cave to evaluate the neural network’s creation. The singer thanked the fan for his attention but criticized the idea of creating songs using artificial intelligence algorithms.
“ChatGPT can write a speech, an essay, a sermon, or an obituary, but it is not capable of creating a real song. Perhaps in time, the algorithm will succeed, and these songs will be indistinguishable from the original ones, but it will still be a copy, a kind of parody. Songs are born out of suffering, they are based on the complex internal creative struggle of a person, but as far as I know, algorithms are not capable of feeling. Data doesn’t suffer,” Nick Cave replied to the fan.
I generally agree with him. An algorithm cannot write a song on the same level as a human, because it requires feelings and personal experience. ChatGPT is still at the initial stage of development, but the point will always remain the same: chatbots will only be able to parody and copy people, but not create something unique.
The full lyrics of the song generated by the neural network can be found on Nick Cave’s website.
Simulate a conversation with a specific person
ChatGPT has the ability to answer calls and have conversations in the way we want. This means that it can explain a certain issue as if it were talking to a child, or it can imitate specific people, historical figures, etc.
Schoolchildren or students will love this useful feature. Imagine this. The law of gravity will be explained to your child not by a boring physics teacher, but by Isaac Newton himself, or rather by his voice imitator.
I would listen to the presentation of the iPhone 15 in the voice of Steve Jobs, and let Elon Musk tell me about the launch of the SpaceX Starship. I like this prospect very much.
Writing phishing emails
Wherever and whenever a new tool appears, there will be attackers who want to use it for their own purposes. The program’s capabilities allow you to create convincing stories in any language, so naturally, people quickly appeared who use this tool to create phishing attacks, fake news, and spread disinformation.
Experts are already warning that Russians want to use Chat GPT for criminal purposes. Check Point Research analysts have discovered underground forums of Russian cybercriminals discussing strategies for the dangerous use of GPT. They are not accessible to the general public, but with proper knowledge, there are no problems with security and geofencing bypass.
Experts express concern about the criminal use of AI. According to Check Point Research, Russians can use it to create fake content and launch phishing campaigns. And they have proof of this.
Chat GPT is a great tool for Russian cybercriminals who don’t speak English. Thanks to quick translations generated by artificial intelligence, they can create reliable phishing emails. And this is just the beginning.
Explaining complex issues
One of the unique features of ChatGPT is its ability to synthesize complex topics and explain them in a simple and logical way. Thanks to this feature, ChatGPT is a great learning tool as it significantly reduces the time required to access key information.
A tool from a powerful AI player called OpenAI allows you to enter questions in any language, to which the chatbot gives conversational, albeit somewhat tricky, answers. The bot memorizes your dialog and uses previous questions and answers to formulate its next responses. The answers are, of course, based on vast amounts of information on the Internet.
The tool appears to be reasonably knowledgeable in areas where there is good underlying data on which it can learn. It is not yet omniscient and intelligent enough to replace all humans, but it can be creative and its answers can sound quite authoritative.
Writing computer code
As it turned out, ChatGPT also allows you to create computer programs. Of course, it quickly became apparent that you can create malware in this way, which is probably not surprising. However, as it turned out, you can also create more complex programs, but with… different results.
There are reports on hacker forums about the first successful attempts to create hacker tools with ChatGPT.
For example, one of the programs was written in Python and, if properly developed, can be quickly turned into a ransomware. The program can encrypt all files on the victim’s computer and demand a fee for decryption.
Another example is that a more advanced hacker posted an example of code written by ChatGPT on a darknet forum that makes a hidden download to the client’s server. By connecting to the server unnoticed, the cybercriminal gets full access to the victim’s data.
But that’s not all. Check Point Research specialists managed to get a whole bunch of tools with ChatGPT: port scanning scripts, access to the Reverse Shell, and even the ability to compile the code into an executable Windows file.
Artificial intelligence is changing almost everything
It’s hard to find any aspect of a person’s mental work that wouldn’t be affected by the amazing capabilities of ChataGPT.
OpenAI writes excellent resumes for job seekers. ChatGPT creates great recipes based on food ingredients we have at home. The AI also creates great investment strategies based on the user’s wealth status and structure.
ChatGPT even passed the American medical exam. Axios researchers who conducted the experiment report that the answers provided by the algorithm were consistent, thorough, and comprehensive. ChatGPT easily passed the exam at one of the US business schools. I am sure that there will be even more stories like this in the future.
OpenAI is also transforming the journalism industry. One of the most popular American new media services, Buzzfeed, announced that it would use ChataGPT to create quizzes and some journalistic content. After this announcement, the stock market price of the media company that publishes Buzzfeed increased by 62%.
We would like to remind you that ChatGPT is only 2 months old. It’s hard to imagine what skills it will have, say, in 2 years.
Are we really facing a revolution? Artists are protesting, teachers and professors are complaining and at the same time afraid, investment advisors are crying, but they will not be able to stop the development of AI in the form of ChatGPT. In the market, you can feel the expectation of powerful shocks, changes that can affect absolutely every aspect of the online world. Something similar to what happened after the launch of the iPhone, but it is still far from the explosion of the popularity of mobile Internet and mobile apps. We are in for some very interesting, busy times. Let’s hope that this is not a terrible version of SkyNet, but something that will help humanity develop, learn and explore the world.
Read also:
- What is a deepfake, why is it dangerous and how to recognize it
- Twitter in Elon Musk’s Hands — A Threat or an “Improvement”?