MIT lists over 700 risks AI poses to humanity; see the 5 worst

The increasing presence of artificial intelligence tools in our daily lives has brought with it a series of discussions about the risks of technology for humanity. One of the latest studies in this direction has been developed by Massachusetts Institute of Technology (MIT) FutureTech groupIn the United States.

The team was formed in collaboration with experts from universities in the US and Australia. with new database 700+ Potential AI RisksThey analyzed various scenarios where smart systems could malfunction and cause different types of damage.

The “Artificial Intelligence Risk Pool,” as the list is called, is being touted as the most comprehensive platform on the subject to date and will continue to be so in the future. support future research into the effects of technologyThe researchers divided the threats into seven domains and 23 subdomains.

According to the authors of the database, the most frequently repeated dangers associated with the use of AI are those related to artificial intelligence. security, bias and discrimination, and privacy issues.

Most of these occurred after the technology was trained and deployed (65%), while only 10% occurred during development.

The 5 biggest risks of artificial intelligence

What could be the biggest dangers of overexposure to intelligent systems? Website based on MIT database Euronews created a list Five main risks of AI in case of failure and harm to humanity.

Check out the top threats by publication:

• Distortion of reality: malicious use of technology can produce content that can spread misinformation and manipulate public opinion;
• Emotional attachment: the human-like language and constant interaction in some bots can lead people to become emotionally attached to the technology;
• People decide: Over-dependence on AIs causes the individual to lose their autonomy by reducing their capacity for critical thinking and problem-solving skills;
• Independence: experts suggest that advanced systems could surpass human intelligence and act autonomously, resisting attempts at control;
• Emotional AI: Another risk outlined in the report is achieving sentience; advanced research could potentially enable AI to feel emotions or sensations such as pleasure and pain.