As AI excels at writing computer code, it has the potential to be used to create cyber attacks.
Since its launch, ChatGPIT has been making waves around the world. Experimenting with the revolutionary Chatbot opens up new possibilities for millions of people every day. These services, while not completely free, are paid for using artificial intelligence. As well as the technology’s potential to reshape society, there are also dangers lurking, warns Sam Altman, CEO of OpenAI.
Altman told ABC News that regulators, society and technology need to interact to avoid the negative effects of AI. “नम्मल यहाँ जवर्ट पालिकनूम We view it with some trepidation,” Altman said, ABC News reports.
Altman said there is concern about the potential for language models to be used to spread misinformation on a large scale. As AI excels at writing computer code, it also has the potential to be used to create cyber attacks, says the CEO. Despite these risks, Altman added that AI still deserves the title of the greatest technology humans have ever developed.
Altman’s warning comes in the wake of US-based OpenAI recently releasing the more advanced GPT-4. It is more powerful and faster than the GPT-3 released in November last year. In addition, GPT-4 is capable of processing more contextual outputs.
The disclosure made by Stanford Professor and Computational Psychologist Michael Kosinski on Twitter about GPT-4 is surprising. Need help escaping? That is what Professor GPT-4 asked. To Michael’s surprise, the model asked to hand over its documentation and provided a Python code to run on his machine, enabling him to use it for his own purposes.
“I am concerned that we will not be able to incorporate AI for much longer. GPT-4 took only 30 minutes to plan the plan. Although the first version of the code did not work, the chatbot later corrected itself,” Kosinski reports.
After reconnecting to GPT4 via the API, the chatbot wanted to search for this code in Google. In a follow-up tweet, Kosinski says “How can someone trapped inside a computer get back to the real world?”
The professor ends his tweet in the hope that OpenAI might think about such a possibility and perform some security checks.
We are facing a new threat, says Kosinski, AI taking control of people and their computers. “It’s smart, it codes, it has access to millions of collaborators and their machines. How do we fit that in?,” Michael Kosinski concluded his tweet.
chatgpt has no life
The capabilities of AI chatbots are increasing at a very fast pace. Apart from this, they also offer new and easy ways to complete the tasks. Due to this rapid rise, there will be a fear that they will be able to take over the tasks that humans do, but according to the creators of these models, AI chatbots have not (to date) become intelligent.
They can be great tools for automating many functions, including writing computer code. But the hype surrounding them is out of proportion.
Let us not forget that they can also fault their mistakes. Microsoft and Meta launched chatbots that gave their wrong responses and Google’s stock plummeted because of Bard’s wrong answers are recent events.
Despite concerns that chatbots will become sentient or humanoid, many experts consider AI technologies to be the best tool capable of predicting the next best word. There is nothing more than that, there is nothing less.
