Artificial intelligence is now a technology that penetrates into our daily lives. A learning model of artificial intelligence, which is used in many areas from phones to home appliances, özgü been developed. This situation is interesting as well as it poses a great danger in case of misuse.
OpenAI was founded as a nonprofit company that will continue its research with the investments of wealthy businessmen. The research team working at the company had explained that an artificial intelligence they had previously created could write a composition. This composition was written on a topic on which the research group disagreed and was detailed enough to convince the team about the subject. This development signaled the team that artificial intelligence would advance rapidly, but even they did not anticipate that an artificial intelligence could do every job in many different areas at the same time.
The mentioned artificial intelligence özgü an algorithm called GPT-2. When this algorithm was first created, it was programmed to perform some tasks. So far, everything may seem düzgüsel, because the situation does not differ much from the AI news you hear often. What happened after this point astonishes not only you, but also the team that created the artificial intelligence.
According to the statements of the research group, he writes such a detailed and complete composition on a subject given to GPT-2 that he can get full marks in SATs (university entrance exams in the USA). Artificial intelligence creates new realities by creating imaginary characters and plots when presented with a story that doesn’t actually exist.
The GPT-2’s tricks don’t stop there. Do you remember the gap-filling questions asked in exams? Often such questions have a specific answer, but the algorithm scans such a large body of knowledge that it can alternatively generate a large number of answers. This feature, which goes far beyond the automatic word completion feature used on phones, offers not only words but also different stories. Somewhat terrifying of the situation is that some of the answers aren’t even thought of by people. The algorithm can also instantly translate works written in many different languages and write texts using it.
Are the distress bells ringing?
What özgü been written so far may seem like heralding the great benefits the advancement of technology will bring to us, but the creators of the algorithm are a little worried about this. The reason for this fear is that the GPT-2 is not actually programmed to do these things. In other words, artificial intelligence can learn on its own. For this reason, the company, which keeps the written codes secret, says that such technology can pose a great danger all over the world if used maliciously.
While artificial intelligence will make our lives unpredictably easier, it also brings dangers with it. What do you think about this subject? Should artificial intelligence continue to be developed? Or should this situation have a limit?
Source: https://www.theverge.com/2019/2/14/18224704/ai-machine-learning-language-models-read-write-openai-gpt2
Dikkat: Sitemiz herkese açık bir platform olduğundan, çox fazla kişi paylaşım yapmaktadır. Sitenizden izinsiz paylaşım yapılması durumunda iletişim bölümünden bildirmeniz yeterlidir.
Source: https://kimdeyir.com/openaiin-yapay-zekasi-kendini-gelistirmeye-basladi/
