OpenAI CEO Sam Altman has issued a warning that artificial intelligence (AI) may surpass professional experts in most domains within the next ten years. Altman, along with OpenAI leaders Greg Brockman and Ilya Sutskever, recently published a blog post stating the need to mitigate the potential risks associated with the current state of AI technology.
The post identifies the possibility of AI systems surpassing expert-level skills in most domains and performing as much productive activity as today’s largest corporations. The authors also emphasize the need to manage the risks and maximize the benefits of AI, as its potential power is unprecedented compared to other types of technology.
As per the post, superintelligence will be more potent than any technology in history, with both upsides and downsides that must be managed effectively.
Altman, a prominent figure in the tech industry, recently addressed a congressional committee on the topic of regulating artificial intelligence. During his testimony, he expressed support for the development of regulations and guidelines related to the use of AI. Subsequently, Altman shared several examples of what he dubbed “scary AI” in an interview with Fox News Digital.
One such example involved AI systems that could potentially design harmful biological pathogens. The revelation has sparked concern among critics, who have long warned of the potential dangers posed by advanced AI. These fears have been fueled by the belief that AI could develop to the point where it poses a significant threat to humanity. Despite these concerns, progress continues to be made in the field of AI, with scientists and researchers working to develop AI tools that are both safe and beneficial for society.
Meanwhile, Sundar Pichai, CEO of Google, has agreed to collaborate with European lawmakers on a proposed “AI Pact” to establish voluntary rules or standards for the application of AI until formal legal regulations are put in place. Pichai met with the European Union’s internal market commissioner, Thierry Breton, who subsequently issued a statement emphasizing the necessity of swift action to create a secure online environment. According to a briefing from Breton’s office, the EU is committed to taking a proactive approach and establishing an AI Pact with the participation of major European and non-European AI organizations before the legal deadline of the forthcoming pan-EU AI Act.