The Year of Superintelligence?

Metatrend #2: AI & Quantum

What it is

"I think we're quite close to digital superintelligence (ASI). It may happen this year, maybe it doesn't happen this year, next year for sure." – Elon

So, what is ASI?

ChatGPT defines ASI as: "An intelligence far surpassing the best human brains in every field—creativity, problem-solving, decision-making—possessing self-improvement, strategic foresight, and vast knowledge beyond human capacity."

Grok-3 defines ASI as: "AI surpassing human intelligence in all intellectual tasks."

My favorite definition is: "Smarter than any human at anything."

Why it matters

There are two major implications of ASI that I care about most:

Implication #1 – “Warp Speed for Science”

With the birth of ASI, we are likely to accelerate scientific discovery 100x in math, physics, chemistry, material sciences, and biology.

This is why Anthropic CEO Dario Amodei, PhD said we could even see a doubling of the human lifespan in the next 5 – 10 years. We could create new catalysts that scrub CO2 from the atmosphere, rapidly desalinate sea water, or double the efficiency of solar. 

How do we think about having AI progeny that may soon be one-billion-fold more intelligent than us? Which, by the way, is numerically equivalent to the intellectual difference between a hamster and a human. With that kind of raw power and intelligence, we can anticipate that AIs will discover many new breakthroughs in physics and realize ingenious solutions to problems like famine, poverty, the climate crisis, and human mortality. 

Implication #2 – “The Last ASI”

Jeff Clune, a DeepMind advisor, warns that "The first ASI is likely to be the last ASI."

He predicts the first ASI could suppress other ASI attempts and trigger a winner-take-all endgame. He also predicts the nationalization of the ASI. As he puts it: "If you’re the prime minister or head of state of a country and somebody, a company within your borders, creates a super weapon, a superpower that is effectively a God, do you nationalize that?" 

Because self-improving AI expands at a super-exponential rate, first place becomes everything.

Have You Heard About Ilya Sutskever's $32B AI Startup?

In June 2024, Ilya Sutskever, PhD founded Safe Superintelligence (SSI) after leaving OpenAI. Ilya’s mission? Develop a fully aligned digital superintelligence.

Incredibly, Ilya was able to raise $6B at a $32B valuation for a company with no known product or revenues. How did he do that? In fact, how did he get the most extraordinary venture investors on the planet to put in the money? The group of VCs includes the following name brands: a16z, Sequoia Capital, DST Global (Yuri Milner), Alphabet, Nvidia, and Lightspeed Venture Partners.

My Answer: I imagine that Ilya was able to convince these investors that he was going to be the very first person to build a digital superintelligence, and that his superintelligence would be the last one and it would be aligned for the benefit of humanity. 

So, if you’re one of these venture funds and you believe him, you would absolutely have to invest based upon Jeff Clune’s thesis that the first ASI is the last ASI. 

The Stakes Have Never Been Higher

As entrepreneurs, innovators, or just plain citizens, we need to be asking the tough questions.

How do we ensure this leap forward benefits humanity? Can we create safeguards? Or are we unleashing forces beyond our control?

Other Key Tech Developments This Week: 

1

Meta creates Superintelligence Labs led by Scale AI and GitHub ex-CEOs as Zuckerberg escalates AI competition

2

Tesla achieves first fully autonomous car delivery from Austin factory to customer without human driver

3

Microsoft pursues medical superintelligence with AI that diagnoses complex cases at 85% accuracy versus 20% for doctors