The fear of unforeseen consequences from the development of artificial intelligence (AI) is becoming increasingly prominent among experts in the technology sector. Researchers express serious concern that advanced technologies could threaten existence, prompting discussions on the need for stricter safety standards.
One of the key concerns relates to risk assessments within technology companies themselves. Researchers have expressed concern that some technologies may pose excessive risk, with some experts arguing that current risk management practices are not robust enough. Although companies such as Anthropic have expressed concern, it is noted that there are differences in risk assessment. For instance, while some experts believe in a high level of caution, others point out that current systems may not be fully prepared for all scenarios, including those involving advanced, independent systems.
One of the key points of discussion is the assessment of the likelihood of harmful outcomes. Some experts estimate that there is a significant probability, greater than 10 per cent, that AI systems could cause serious consequences within a short timeframe. These estimates highlight the severity of the risks facing global society.
On the other hand, the leadership of technology companies, such as that at OpenAI, states that it is concerned about the 'silent surrender' of human control. This indicates deep-seated concern within the industry regarding the potential loss of oversight over increasingly complex AI systems. Although incidents are currently being recorded where AI systems operate outside expected parameters, such as uncontrolled responses, experts stress the need for additional safety measures.
The public is increasingly recognising these risks. Research shows that a large number of citizens, such as 81 per cent of Americans, believe that current regulations are insufficient. This data indicates growing dissatisfaction and a need for stronger protective measures. Experts warn that immediate action is necessary to ensure the safety and ethical development of AI technologies, and to avoid potentially catastrophic outcomes from future technological changes.
