Sam Altman: Superintelligence a Few Thousand Days Away
Last month, OpenAI CEO Sam Altman offered an even bolder prediction: the superintelligent AI could happen in a few thousand days. This has been celebrated online by Altman in a blog post, which he describes as the “Intelligence Age.
AI MegaSoar: A Singularity Milestone
As Altman points out, the advent of superintelligent AI, i.e., an AI that has the cognitive capacities to exceed those of humans, might constitute one of the most significant events in human history. Unlike existing AI models like ChatGPT that need thousands of words worth of writing data to learn from, superintelligence could teach itself and iterate new versions without human intervention.
It is possible we may have superintelligence a few thousand days hence (! ); it will take a bit more time but I know we can get there.
The Need for Infrastructure
Superintelligence costs tremendous amounts of computational resources and infrastructure, he stated in his blog post. He emphasized the inclusion of Artificial Intelligence within everyone’s reach and not just in the hands of the rich and famous.
“If there’s not enough infrastructure, AIG will be a scarce resource that people go to war over and only rich corporations get to use,” Altman said.
Energy and Chips are the critical where all this revolution in AI is revolving around. Altman demanded that the price of computing power — vital for AI capabilities — be significantly lowered to afford all players alike.
Making the Best of the Risks of SuperIntelligence
Altman agreed with the potential value of superintelligence, but also noted its hazards. In the end, he warned humanity to use this monumental technology with caution, knowing its potential dangers and benefits.
Avi wrote, “The story will not be all happy, but the potential benefit is so great that we owe it to ourselves — and to our future — to learn how best to steer through the dangers ahead of us.”
A Brand New Startup About Safe Superintelligence
Relatedly, a new startup, Safe Superintelligence (SSI), from former OpenAI chief scientist Ilya Sutskever in the past month has become unicorn with $1B to develop a safe and reliable superintelligent AI. Sutskever has combined with co-founders Daniel Levy and new Apple AI chief Daniel Gross using the objective to make advances in the field of AI without having mental distribution in order to manage products along with product cycles.
Gross told Reuters, “It is critical for us to be surrounded by investors who realize the significance of our mission — that of creating a safe path to superintelligence.
Conclusion
Sam Altman and other industry titans are working to harness superintelligent AI, cracking the toughest engineering challenges along the way and tackling a thornier challenge: ensuring AI has the moral compass it needs highbe.forbes.com The road to super-intelligence may be much shorter than we think Published at May 21, 2015 by Sam Altman You needn’t understand all of that Andreev or Turing stuff juicy details on edge-called self-improvement here that you accused me of holding back-of your time. Under the correct infrastructure and learning based approach, superintelligence could transform the world like nothing we have ever seen — offering tremendous opportunities as well as dire challenges.