As the pace of technological advancements accelerates, the advent of Artificial General Intelligence (AGI) draws nearer – making it essential that we proactively prepare for its implications.

Notable experts in AI, including Anthropic’s Dario Amodei and OpenAI’s Sam Altman, caution that highly effective AI or even superintelligence could potentially emerge within the next two to ten years, with far-reaching implications for our world.

Amodei’s essay offers a thoughtful examination of AI’s capabilities, hypothesizing that the development of highly advanced AI – commonly referred to as synthetic general intelligence (AGI) – may become a reality by as early as 2026. In the near future, according to Altman, it’s possible that humanity will be on the cusp of developing superintelligence within a mere few thousand days – potentially by as early as 2034. By some estimates, a revolution is poised to reshape the global landscape over the next decade, with potentially transformative consequences unfolding within the next 2-10 years.

Amodei and Altman, pioneers in AI analysis and improvement, drive innovation by expanding the limits of what’s possible, thereby shaping the future with their impactful findings. Altman’s essay implies the concept of superintelligence as artificial intelligence systems that outperform human cognition across various disciplines, including biology, programming, mathematics and engineering, mirroring Amodei’s definition of “smarter than a Nobel Prize winner throughout most related fields”. 

While not all share the optimism, many tech leaders remain undeterred by more pessimistic views. Ilya Sutskever, a co-founder of OpenAI, has recently founded Protected Superintelligence (SSI), a startup committed to developing AI with a prioritization on safety from the outset. As Sutskever concluded the SSI final in June, he proclaimed, “We will relentlessly pursue protected superintelligence with unwavering focus, singular aim, and a single-minded product.” Later, at OpenAI, he predicted, “This endeavour will have a monumental, earth-shaking impact. Given the early stages of this research capability, Sutskever is poised to make significant strides in funding innovative firm initiatives ahead of and subsequent to its launch at SSI.

By some estimates, AI is expected to surpass human capabilities by 2029, a notion shared by entrepreneur Elon Musk. By Musk’s prediction, AI will possess capabilities rivalling those of humans within the next year or two. By 2028 or 2029, he predicted that AI will be able to accomplish what a group of humans working together could achieve in just an additional three years. According to the longstanding views of renowned futurist Ray Kurzweil, the potential emergence of AGI as early as 2029 is a prediction consistent with his outlook on artificial intelligence’s trajectory. In his seminal work published in 1995, Raymond Kurzweil predicted the future with uncanny accuracy. His bestselling e-book, “The Singularity is Near”, released in 2005, foresaw many of the technological advancements that have since become a reality. 

The upcoming transformation

As we stand at the threshold of imminent breakthroughs, it’s crucial that we critically evaluate our readiness to absorb and implement their transformative impact. Regardless of preparation, if these prophecies prove accurate, a fundamentally novel world will rapidly emerge. 

As we converse, a newborn may enroll in kindergarten in a world transformed by the advent of Artificial General Intelligence (AGI). As robots become increasingly sophisticated, the prospect of AI-powered caregivers gaining traction in the near future seems inevitable. As society hurtles toward a future where artificial intelligence and robotics are increasingly intertwined with human life, the notion of an android companion for teenagers, as envisioned by Kazuo Ishiguro in “Never Let Me Go”, no longer seems entirely implausible. The prospect of AI companions and caregivers hints at a world poised to undergo profound moral and societal upheavals, potentially challenging the very foundations of our current frameworks.

Can long-standing relationships between humans and technology foster a profound synergy, or will the consequences of such an alliance precipitate catastrophic uncertainty? The potential benefits that could arise from. With future robotic breakthroughs, humanity may potentially unlock cures for the majority of cancers and depression, ultimately achieving a long-sought milestone in fusion energy. Some envision this forthcoming era as a time when people will have novel opportunities for creativity and collaboration. Despite the optimistic outlook, the credible drawbacks are equally significant, encompassing massive joblessness, stark revenue disparities, and uncontrolled autonomous weaponry. 

Within a relatively short timeframe, MIT Sloan principal analytics scientist Andrew McAfee views AI as augmenting rather than replacing human jobs. Currently, he posits that AI furnishes a veritable army of clerks, colleagues, and coaches at one’s disposal, with the added benefit that it often assumes responsibility for “large swaths” of tasks. 

However, this measured view of AI’s impact may have a limited shelf life. Elon Musk posits that, in the long run, “most likely none of us can have a job” – a sobering assertion that underscores an essential point: regardless of how true AI’s capabilities and impacts may appear in 2024, they could be radically altered within the AGI world just a few years away.

Tempering Expectations: The Art of Harmonizing Optimism and Reality

Despite bold predictions, not everyone concurs that highly effective AI is imminent or that its outcomes will be straightforward. Renowned sceptic Gary Marcus has been cautioning that current AI technologies are unlikely to yield genuine Artificial General Intelligence (AGI), contending that the field lacks the requisite capacity for deep, nuanced reasoning. Famed expert Eliezer Yudkowsky publicly challenged Elon Musk’s assertion that AI would soon surpass human intelligence, wagering a million dollars on his prediction being inaccurate.

Linus Torvalds, the creator and lead developer of the Linux operating system, has expressed skepticism about artificial intelligence (AI), describing it as “90% advertising and marketing” with only 10% actual substance. He suggests that at present, AI is more hype than reality.

Perhaps bolstering Torvald’s claim is an OpenAI revelation showing their flagship LLM models, GPT-4o and GPT-o1, stalling on straightforward questions with verifiable answers. While the paper introduces a novel “SimpleQA” benchmark to evaluate the veracity of linguistic models, it is striking that even the top-performing model, o1-preview, exhibited poor accuracy, failing to provide correct answers for roughly 50% of the queries. 

Futuring Forward: Preparing for the Age of Artificial Intelligence

The vast promise of AI lies in its burgeoning capabilities, as exemplified by recent strides in benchmarking platforms such as SimpleQA, which have consistently demonstrated the technology’s remarkable aptitude for discernment and insight. While significant progress has been made in AI research, substantial advancements are still required to achieve true Artificial General Intelligence (AGI). 

Despite this, those most familiar with AI’s creation predict rapid progress nonetheless. According to OpenAI’s former senior adviser on AGI readiness, Martine Brundage, there is a consensus among experts that Artificial General Intelligence (AGI) will emerge relatively soon. However, she cautions that the implications of this development for society are impossible to predict with certainty.

Stanford’s Roy Amara coined the concept of Amara’s Regulation in 1973, noting that people often overestimate the short-term impact of emerging technologies and underestimate their long-term implications. While the exact timing of AGI’s arrival may deviate from the most ambitious forecasts, its impending emergence, potentially within a few years, will likely have a profound impact on society that surpasses even today’s most optimistic projections. 

Despite progress made in AI capabilities, a vast chasm still separates them from true Artificial General Intelligence. The stakes involved in this scenario are high, ranging from groundbreaking medical advancements to catastrophic threats to humanity’s very existence. It’s imperative that we establish robust security frameworks, retool our organizations, and prepare for a transformation that will fundamentally change the human experience? The real concern is not just when Artificial General Intelligence (AGI) will emerge, but also whether we’ll be adequately prepared for its arrival when it does.