AI Singularity Could Be Here Soon

The Singularity: Just Months Away?

Since large language models (LLMs) have become widespread, opinions on AI's future are everywhere. Some researchers believe the singularity, where machines surpass human intelligence, might be decades away. However, the CEO of Anthropic thinks it could be as close as three months.

Analyzing AI Predictions

A new analysis from AIMultiple examined 8,590 predictions from scientists, entrepreneurs, and others to understand the varying AI forecasts. The study tracked changes in these predictions over time using data analysis techniques. The research looked at different AI thresholds, including artificial general intelligence (AGI) and AI superintelligence.

AI industry leaders were generally more optimistic in their predictions. Most believed AGI would likely occur within the next half-century. The arrival of LLMs has changed timelines for AGI and the singularity. "Current surveys of AI researchers are predicting AGI around 2040," the report notes. "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."

Why AGI Seems Inevitable

The analysis provides insight into why many experts think AGI is coming. Machine intelligence, unlike human intelligence, appears limitless. Computing power doubles roughly every 18 months, potentially allowing LLMs to match human calculation speeds. If computing faces limitations, quantum computing could compensate. The report mentions, "Most experts believe that Moore’s law is coming to an end during this decade. 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."

Is AGI a Certainty?

However, not everyone is convinced AGI is guaranteed. Some argue that human intelligence is more complex than current AGI definitions suggest. Some AI experts view the human mind as having multiple intelligences, including interpersonal, intrapersonal, and existential intelligence, in addition to logical-mathematical intelligence. Deep learning pioneer Yann LeCun suggests renaming AGI to “advanced machine intelligence,” arguing that human intelligence is too specialized to be replicated.

The report also implies that AI can aid in discoveries, but not independently. “More intelligence can lead to better-designed and managed experiments, enabling more discovery per experiment,” the report states. “Even the best machine analyzing existing data may not be able to find a cure for cancer.”

Although AGI predictions vary, it's clear that algorithms will bring significant change. The impact of these changes is up to us.