Alan Turing Institute Study: Risk of Losing Control of AI in the Next Five Years
Experts warn of the emergence of “super-intelligent” systems that are superior to humans in cognitive tasks
Quick Look
- Research from the Alan Turing Institute points to the risk of losing control of AI within the next five years.
- The emergence of “superintelligent” systems that are superior to humans in cognitive tasks may force people to delegate critical decisions to them.
AI-generated summary
Why It Matters
A study from the Alan Turing Institute analyzes the risks associated with the pace of development of artificial intelligence.
MOSCOW, September 23 - RIA Novosti. People may lose control over artificial intelligence systems over the next five years due to the rapid development of such systems, says a study by the Alan Turing Institute in Britain.
“There is a realistic possibility that within the next five years, ‘superintelligent’ AI systems will emerge that will outperform humans at virtually all important cognitive tasks,” the study says.
It is noted that the speed and quality of task performance by AI systems may increasingly force people to delegate decision-making to them.
“If AI systems invent strategies that institutions cannot adequately evaluate, and also work faster than organizations, significant human control over them may disappear,” the report says.
Open Questions
- Which specific institutions are at risk of losing control?
- What security measures are offered to prevent loss of control?
