Peter H. Diamandis BLOG - Upgrade Your Mindset.

120 IQ AI: Threat or Opportunity?

Written by Peter H. Diamandis | Sep 26, 2024

IQ Unmasked: Are You Smarter Than an AI?

 

To understand the implications of this breakthrough, it helps to recap what exactly IQ is. According to Mensa, the largest and oldest high-IQ society in the world:

 

The term “IQ” was coined in 1912 by psychologist William Stern. Originally, IQ was represented as a ratio of mental age to chronological age x 100. Today, IQ tests are designed so that the average score of the norming sample is defined as IQ 100, with standard deviations up or down.

 

Mensa admits individuals who score in the top 2% of the population, which means scoring 132 or more in the Stanford-Binet test, or 148 or more in the Cattell equivalent.

 

As an article in Vox points out, most people have average intelligence. IQ is a composite score made up of many different tests of reasoning, memory, acquired knowledge, and mental processing speed. These sub-scores are totaled and compared with those of the rest of the population.

 

A perfectly average score is set at 100. How does this chart of Global Average IQ scores make you feel?

 

 

Most people you meet are probably average, and a few are extraordinarily smart. Just 2.2 percent have an IQ of 130 or greater.

 

"The classic finding — I would say it is the most replicated finding in psychology — is that people who are good at one type of mental task tend to be good at them all," says Stuart Ritchie, an intelligence researcher at the University of Edinburgh.

 

This is why OpenAI's model o1 achieving an IQ score of 120 is so significant. It's not just outperforming the average human—it's surpassing 91% of the population in a test designed to measure general intelligence.

 

AI Showdown: o1 Leaves Claude, Gemini & Grok Behind

 

Maxim Lott, creator of Tracking AI which monitors bias in AI chatbots, writes on his site maximumtruth.org, "Worried about AI taking over the world? You probably should be." 

 

Lott posted the below image to X last week, and it's been making the rounds:

 

 

As Lott puts it, "OpenAI's new o1 model is a BIG breakthrough in AI intelligence, if IQ tests say anything. I gave it the Norway Mensa IQ test, and it blows other AIs out of the water. I'm surprised!... Because there hadn't been public progress in the last 6mo."

 

Lott continues, "I had become blasé about AI progress after my initial tests in February, because there was approximately zero IQ improvement since then. This week, that all changed. Specifically, o1 got 25 out of 35 IQ questions correct, far above what most humans get."

 

Importantly, Lott points out that this improvement isn't because the questions are in the training data. He created a new, offline-only test to ensure the questions weren't accessible to AI training data. The results? O1 still showed a massive improvement in reasoning ability.

 

The AI IQ Race: Buckle Up for Exponential Growth

 

Lott's analysis of Anthropic's Claude model progression is enlightening:

  • Claude-2 (July 2023, 4 months later): ~82 IQ
  • Claude-3 (March 2024, 8 months later): ~101 IQ

 

Based on this trend, Lott projects:

  • Claude-4 (expected in 12-16 months): ~120 IQ
  • Claude-5 (16-32 months after Claude-4): ~140 IQ
  • Claude-6 (20-64 months after Claude-5): Smarter than everyone

Lott notes:

 

"I now think that timeline is holding up, and that OpenAI has always been about 6 months ahead of Anthropic/Claude, behind the scenes. If so, then we should start seeing AIs breaking 140 IQ in 2026."

 

Countdown to Superintelligence: Brace for Impact

 

As CEO of OpenAI, Sam Altman knows more about the current capabilities of AI than almost anyone else in the world. In his recent article The Intelligence Age, he predicts that we'll see digital superintelligence "in a few thousand days."

 

What's driving this progress? Altman explains:

  • "In three words: deep learning worked”.
  • “In 15 words: deep learning worked, got predictably better with scale, and we dedicated increasing resources to it.”

That's really it; humanity discovered an algorithm that could really, truly learn any distribution of data (or really, the underlying "rules" that produce any distribution of data). To a shocking degree of precision, the more compute and data available, the better it gets at helping people solve hard problems."

 

Implications of the AI Revolution: Your World is About to Change

 

The results of OpenAI's o1 model suggest AIs are developing true intelligence that could have profound implications in the coming years.

 

So, what happens when AIs are smarter than humans? What will we be able to accomplish?

 

According to Sam Altman:

  1. We'll soon be able to work with AI that helps us accomplish much more than we ever could without AI.

  2. Eventually, we can each have a personal AI team, full of virtual experts in different areas.

  3. AI models will serve as autonomous personal assistants who carry out specific tasks on our behalf.

  4. Our children will have virtual tutors providing personalized instruction in any subject.

  5. We can expect better healthcare, the ability to create any kind of software imaginable, and much more.

  6. AI systems will help us make better next-generation systems and accelerate scientific progress across the board.

Altman believes this will lead to unprecedented prosperity:

 

"I believe the future is going to be so bright that no one can do it justice by trying to write about it now; a defining characteristic of the Intelligence Age will be massive prosperity."

 

As entrepreneurs, business leaders, and individuals, we need to start preparing for this new era. How will you leverage these AI advancements in your business? How will you adapt your skills and knowledge to stay relevant in an AI-driven world?

 

The future is approaching faster than we might think. It's time to embrace the potential of AI and start envisioning how we can use it to create a better world for all of us.