Human-level AI could be here as early as 2026: Anthropic CEO

As a seasoned researcher with over two decades of experience in the field of artificial intelligence, I must admit that the rapid advancements we are witnessing today leave me both exhilarated and cautious. Having seen the evolution of AI from its infancy to the sophisticated systems we have now, I find it hard not to be amazed by the progress made in such a short span of time. However, my years of study and observation have also instilled in me a healthy dose of skepticism and caution.


According to the CEO of Anthropic, Dario Amodei, we might witness artificial intelligence reaching parity with human-level intelligence by as early as 2026.

This past week, Anthropic CEO Dario Amodei likened the current advancements in Artificial General Intelligence (AGI) to various stages of an educational journey.

He mentioned that our progress has advanced significantly, from being similar to an undergraduate student a few years ago, then moving up to a PhD-level recently. He further speculated that AI might surpass human abilities in just a few more years.

“If you just eyeball the rate at which these capabilities are increasing, it does make you think that we’ll get there by 2026 or 2027.” 

He noted that various factors might halt these progressions, for instance, exhaustion of data resources, constraints in clustering expansion, and international disputes affecting the manufacture of microchips.

Given the projected pace of progress, it’s estimated we may reach our goal sometime between 2026 and 2027. However, there could be a slight postponement from this timeline, according to his statement.

Human-level AI could be here as early as 2026: Anthropic CEO

Additionally, Amodei voiced apprehension regarding the possible consequences of advanced AI systems, remarking, “Powerful tools can be used for beneficial purposes, but they can also cause harm.” Following this, he quoted a well-known phrase: “With great power comes great accountability.

Among various chatbot models, Anthropic’s standout product, Claude, is a significant competitor to ChatGPT from OpenAI, demonstrating similar capabilities.

Speaking about rivalry, he emphasized that the company’s objective was to strive for excellence, aiming to motivate others to act ethically by demonstrating positive behavior ourselves.

When Fridman inquired about the release date for the latest version of the company’s AI chatbot, Claude Opus 3.5, Amodei declined to provide a specific timeline.

According to recent reports, it’s been suggested by OpenAI CEO Sam Altman that the company could potentially reach Artificial General Intelligence (AGI) within the next five years, using existing hardware. This is a prediction similar to those he has made in the past regarding the arrival of AGI within a year or two.

Artificial General Intelligence (AGI) refers to a form of artificial intelligence that can perform tasks and think like a human, demonstrating comparable abilities across various mental functions.

Read More

2024-11-14 08:14