The Singularity Is Nearer: When We Merge with AI

by Ray Kurzweil

Increasing material prosperity has a mutually reinforcing relationship with declining violence. People with a lot to lose economically have stronger incentives to avoid fighting, and when people can look forward to long lives of safety, they have good reason to make long-term investments that benefit society.

The Snapdragon 810, a common chip on smartphones in the $50 range, averages about three billion floating-point operations per second (GFLOPS) across a range of performance benchmarks.[203] This corresponds to around 60 million computations per second per dollar. The best available computers in 1965 achieved about 1.8 computations per second per dollar, and in 1985 they were up to about 220.[204] At those efficiencies, it would have taken almost $1.7 billion (in 2023 dollars) to match the Snapdragon in 1965 and $13.6 million in 1980.

As I have argued in this chapter, contrary to many popular assumptions, life is getting better in profound and fundamental ways for the great majority of people on earth.

Waymo’s self-driving vehicles have traveled well over 20 million fully autonomous miles at the time of this writing

because in many areas AI can actually do a better job than the humans it is replacing. Self-driving cars will be much safer than those operated by human drivers, and the AI will never get drunk, drowsy, or distracted.

just growing and distributing food required nearly all human labor as recently as two centuries ago, while food production in the United States and much of the developed world requires less than 2 percent of labor today.

The US regulatory process involves three main phases of clinical trials, and according to a recent MIT study, only 13.8 percent of candidate drugs make it all the way through to FDA approval.[5] The ultimate result is a process that typically takes a decade to bring a new drug to market, at an average cost estimated between $1.3 billion and $2.6 billion.[6]

In 2020 a team at MIT used AI to develop a powerful antibiotic that kills some of the most dangerous drug-resistant bacteria in existence. Rather than evaluate just a few types of antibiotics, it analyzed 107 million of them in a matter of hours and returned twenty-three potential candidates, highlighting two that appear to be the most effective.[9]

But by far the most important application of AI to medicine in 2020 was the key role it played in designing safe and effective COVID-19 vaccines in record time. On January 11, 2020, Chinese authorities released the virus’s genetic sequence.[11] Moderna scientists got to work with powerful machine-learning tools that analyzed what vaccine would work best against it, and just two days later they had created the sequence for its mRNA vaccine.[12] On February 7 the first clinical batch was produced.

Before the pandemic, vaccines typically took five to ten years to develop.

The 2030s will usher in the third phase of life extension, which will be to use nanotechnology to overcome the limitations of our biological organs altogether. As we enter this phase, we’ll greatly extend our lives, allowing people to greatly transcend the normal human limit of 120 years.[94]

For example, an American man at age 97 has about a 30 percent chance of dying before 98, and if he makes it that far he will have a 32 percent chance of dying before 99. But from age 110 onward, the risk of death rises by about 3.5 percentage points a year.

This is longevity escape velocity.[100] This is why there is sound logic behind Aubrey de Grey’s sensational declaration that the first person to live to 1,000 years has likely already been born.

Recall my estimate that the computation inside the human brain (at the level of neurons) is on the order of 1014 per second. As of 2023, $1,000 of computing power could perform up to 130 trillion computations per second.[130] Based on the 2000–2023 trend, by 2053 about $1,000 of computing power (in 2023 dollars) will be enough to perform around 7 million times as many computations per second as the unenhanced human brain.[131]

Humanity currently has roughly 12,700 nuclear warheads, around 9,440 of which are active and could be used in a nuclear war.[3] The United States and Russia each maintain around 1,000 large warheads that could be launched with less than a half hour’s notice.[4]

As of 2023, there are five nations known to have a full “triad” of nuclear weapons (intercontinental ballistic missiles, air-delivery bombs, and submarine-launched ballistic missiles): the United States (5,244 warheads), Russia (5,889), China (410), Pakistan (170), and India (164).[8] Three other nations are known to have a more limited form of delivery system: France (290), the United Kingdom (225), and North Korea (around 30).