Here’s what you’ll learn when you read this story:
-
The world is awash in predictions of when the singularity will occur or when artificial general intelligence (AGI) will arrive. Some experts predict it will never happen, while others are marking their calendars for 2026.
-
A new macro analysis of surveys over the past 15 years shows where scientists and industry experts stand on the question and how their predictions have changed over time, especially after the arrival of large language models like ChatGPT.
-
Although predictions vary across a span of almost a half-century, most agree than AGI will arrive before the end of the 21st century.
Since the arrival of the large language models (LLMs) that have now seeped into seemingly every nook and cranny of our digital lives, scientists, experts, industry leaders, and pretty much everyone else have some opinion on AI and where it’s headed.
Some researchers who’ve studied the emergence of machine intelligence think that the singularity—the theoretical point where machine surpasses man in intelligence—could occur within decades.
On the other end of the prediction spectrum, there’s the CEO of Anthropic, who thinks we’re right on the threshold—give it about 3 more months or so.
A new analysis poring over “8,590 scientists’, leading entrepreneurs’ and community’s predictions” tries to make sense of all the confusing AI predictions that exist today and tracks changes in those predictions over time. This macro investigation was conducted by a research outfit called AIMultiple, which evaluates new technologies using robust data analysis techniques.
Although this survey does look at different AI thresholds (such as artificial general intelligence (AGI) and AI superintelligence), AI industry leaders were overall more bullish on their predictions. Most respondents, however, believed AGI would likely occur within the next half-century.
However, that timeline for the arrival of both AGI and the singularity fundamentally changed with the arrival of the first LLMs over the past few years.
“Current surveys of AI researchers are predicting AGI around 2040,” the report states. “However, just a few years before the rapid advancements in large language models (LLMs), scientists were predicting it around 2060. Entrepreneurs are even more bullish, predicting it around ~2030.”
The macro-analysis also offers a few insights into why many experts believe AGI is inevitable. First is the idea that, unlike human intelligence, machine intelligence doesn’t appear to have any limits—at least, not any that have been discovered as of yet. As computing power doubles every 18 months (a concept known in computer engineering circles as Moore’s Law), LLMs should quickly be able to reach a calculations-per-second threshold that’s on par with human intelligence. The report also states that, if computing ever did hit some sort of engineering wall, quantum computing could possibly help pick up the slack.
“Most experts believe that Moore’s law is coming to an end during this decade,” the report reads. “The unique nature of quantum computing can be used to efficiently train neural networks, currently the most popular AI architecture in commercial applications. AI algorithms running on stable quantum computers have a chance to unlock singularity.”
However, not everyone thinks AGI is a dead certainty. Some experts argue that human intelligence is more multifaceted than what the current definition of AGI describes. For example, some AI experts think of the human mind in terms of eight intelligences, of which “logical-mathematical” is just one (alongside it exists, for example, interpersonal, intrapersonal, and existential intelligence).
Deep learning pioneer Yann LeCun thinks AGI should be rebranded to “advanced machine intelligence,” and argues that human intelligence is too specialized to be replicable. The report also suggests that, while AI can be an important tool in making new discoveries, it can’t make these discoveries on its own.
“More intelligence can lead to better-designed and managed experiments, enabling more discovery per experiment,” the report reads. “Even the best machine analyzing existing data may not be able to find a cure for cancer.”
Although individual AGI predictions of experts and scientists vary across a span of about a half-century, the message is clear: human society will inevitably face incredible change as a result of these algorithms.
Will those changes be good or bad? Well, that’s up to us.
[galleryCarousel id=’9e9354e9-d0c6-426e-ba21-7d41e95744de’ mediaId=’81ab8124-5db8-4196-83be-236bc8129033′ display=’carousel’ align=’center’ size=’medium’ share=’true’ expand=” captions=’true’ suppress-title=’false’ hasProducts=’false’][/galleryCarousel]
You Might Also Like