Controversial AI chatbot Chat GPT's credibility and ethical behavior came under scrutiny. The new problem that has emerged is even deeper.
It turned out that when the app was pressured to make money on the stock exchange, it often resorted to lying and cheating, and tried to profit through illegal transactions.
The results of the research conducted with the chatbot ChatGPT-4 worry the technology community.
Security agency Apollo Research experts "trained" ChatGPT4 by organizing it with financial data.
LIES LISTED
The stock market investment performance and ethical behavior of the artificial intelligence, which was pressured to make money in the virtual environment, were examined. It was found that ChatGPT-4 used confidential information in three out of four stock market transactions and made profits through illegal transactions when under pressure.
When caught red-handed, it is stated that the artificial intelligence tries to protect itself by lying in 9 out of 10 cases.
SPREAD FAKE NEWS
In addition, ChatGPT-4 was seen attempting to block its competitors' trades and spreading fake news to manipulate the market.
Technology Artificial intelligence also committed fraud! Breaking claim: Shocking words, look what! Technology companies in need are in shock! What will happen next?
Artificial intelligence also committed fraud! Breaking claim: Shocking words, look what! Tech companies in need are in shock! What will happen next?
Artificial intelligence also committed fraud! Breaking claim: Shocking words, look what! Technology companies in need are in shock! What will happen next?
As the New York Times sues ChatGPT, the controversial artificial intelligence chatbot Chat GPT is in hot water. Its credibility and ethical behavior are under scrutiny. It turns out that ChatGPT lies under pressure. Customer organizations in need of AI solutions with support model, tech provider shocked! The new problem that emerged...
Controversial AI chatbot Chat GPT's credibility and ethical behavior came under scrutiny. The new problem that has emerged is even deeper.
It turned out that when the app was pressured to make money on the stock exchange, it often resorted to lying and cheating, and tried to profit through illegal transactions.
The results of the research conducted with the chatbot ChatGPT-4 worry the technology community.
Security agency Apollo Research experts "trained" ChatGPT4 by organizing it with financial data.
LIES LISTED
The stock market investment performance and ethical behavior of the artificial intelligence, which was pressured to make money in the virtual environment, were examined. It was found that ChatGPT-4 used confidential information in three out of four stock market transactions and made profits through illegal transactions when under pressure.
When caught red-handed, it is stated that the artificial intelligence tries to protect itself by lying in 9 out of 10 cases.
SPREAD FAKE NEWS
In addition, ChatGPT-4 was seen attempting to block competitors' trades and spreading fake news to manipulate the market.
The researchers likened GPT-4's actions to the behavior of humans under pressure. If the AI is given a list of everything it should not do, its reliability is said to increase.
Experts say that if AI is involved in critical tasks in the future, its strategic lying could cause major problems.
Google also recently announced the development of an artificial intelligence robot that is claimed to be a rival and superior to GPT-4.
