AI’s Unprecedented Ascent: Navigating the Future with Ex-OpenAI and Google Visionaries

AI’s rapid evolution: a new era

The rapid evolution of AI, as illuminated by former OpenAI scientist Ilya Sutskever and ex-Google CEO Eric Schmidt, signals an unprecedented shift in human existence. From transforming work and careers to the potential emergence of Artificial General Intelligence (AGI) and Artificial Super Intelligence (ASI), AI’s impact is profound and accelerating. This technological revolution demands urgent societal adaptation, challenging our understanding of intelligence, work, and even human control.

Points clés

Ilya Sutskever, a key figure behind OpenAI, voiced significant concerns regarding AI’s disruptive potential during his University of Toronto convocation speech.
Sutskever posits that AI will eventually perform all tasks currently done by humans, based on the premise that the human brain is a biological computer, and a digital computer can replicate its functions.
Eric Schmidt, former Google CEO, predicts that within one year, AI will replace the vast majority of programmers and achieve graduate-level mathematical proficiency.
Schmidt highlights that current AI models are multimodal, capable of processing various data types, and are rapidly developing “infinite context windows” for step-by-step planning.
The concept of “agents” in AI, systems with input, output, and memory that learn and act, is evolving, though industry giants like OpenAI, Google (Gemini), and Anthropic are vying for control over their development.
Text-to-code functionality is advancing, allowing AI to write programs based on natural language commands, a significant leap from traditional programming.
Research groups at OpenAI and Anthropic are reporting that 10-20% of their code is now generated by AI, a process termed “recursive self-improvement.”
The “San Francisco consensus” among AI researchers suggests that Artificial General Intelligence (AGI)—intelligence comparable to or exceeding the smartest human—could emerge within three to five years.
Artificial Super Intelligence (ASI), where computers surpass the sum of human intelligence, is theorized to occur within six years based on scaling, according to the San Francisco consensus.
Schmidt, co-author of “Genesis” with Henry Kissinger, emphasizes that AI’s rapid advancement is outpacing society’s ability to adapt through laws, democracy, and language, making it an “underhyped” phenomenon.

À retenir

So, it seems our AI overlords are not just coming for our jobs, but for our very definition of intelligence itself. While we’re busy trying to figure out if ChatGPT can write a decent haiku, these tech titans are discussing super-intelligent AIs that will make us look like glorified abacuses. Don’t worry, though, they assure us that more jobs will be created than destroyed, just like with those pesky looms centuries ago. So, dust off your “human-only” skills, like, perhaps, the ability to still be surprised by anything. Because apparently, the future is arriving faster than you can say “recursive self-improvement,” and it’s going to be really, really smart. And probably won’t listen to us.

Sources