Like many in Silicon Valley, technology entrepreneur Bryan Johnson sees a future in which intelligent machines can do things like drive cars on their own and anticipate our needs before we ask.
What's uncommon is how Johnson wants to respond: Find a way to supercharge the human brain so that we can keep up with the machines.
From an unassuming office in Venice Beach, his science fiction-meets-science start-up, KerNEL, is building a tiny chip that can be implanted in the brain to help people suffering from neurological damage caused by strokes, Alzheimer's, or concussions.
The team of top neuroscientists building the chip - they call it a neuroprosthetic - hope that in the longer term, it will be able to boost intelligence, memory, and other cognitive tasks.
The medical device is years in the making, Johnson acknowledges, but he can afford the time. He sold his payments company, Braintree, to PayPal for $800 million 2013. A former Mormon raised in Utah, the 38-year-old speaks about the project with missionary-like intensity and focus.
"Human intelligence is landlocked in relationship to artificial intelligence - and the landlock is the degeneration of the body and the brain," he said, in an interview about the company, which had not discussed publicly before.
"This is a question of keeping humans front and center as we progress."
Johnson stands out among an elite set of entrepreneurs who believe Silicon Valley can play a role in funding large-scale scientific discoveries - the kind that can dramatically improve human life in ways that go beyond building software.
Though many of their ventures draw from software principles: In the last two years, venture capital firms like Y-Combinator, Andreessen Horowitz, Peter Thiel's Founder's Fund, Khosla Ventures and others have poured money into start-ups that focus on "bio-hacking" - the notion that you can engineer the body the way you would a software program.
They've funded companies that aim to sequence the bacteria in the gut, reprogram the DNA you were born with, or conduct cancer biopsies from samples of blood.
Human intelligence is landlocked in relationship to artificial intelligence - and the landlock is the degeneration of the body and the brain.
They've backed so-called cognitive enhancement businesses like Thync, which builds a headset that sends mood-altering electrical pulses to the brain, and Nootrobox, a start-up that makes chewable coffee supplements that combine doses of caffeine with active ingredients in green tea, leading to a precisely-engineered, zenlike high.
It's easy to dismiss these efforts as the hubristic, techno-utopian fantasies of a self-involved elite that believes it can defy death and human decline - and in doing so, confer even more advantages on the already-privileged.
And while there's no shortage of hubris in Silicon Valley, it's also undoubtable some of these projects will accelerate scientific breakthroughs and fill some of the gaps left in the wake of declining public funding for scientific research, said Laurie Zoloth, professor of Medical Humanities and Bioethics at Northwestern University.
Moreover, techies are motivated by the fact that many biological and health challenges increasingly involve data-mining and computation; they're looking more like problems that they know how to solve. Large-scale genome sequencing, for example, has long been seen as key to unlocking targeted cancer therapies and detecting disease far earlier than current methods; it's becoming more of a reality as the cost of sequencing, storing, and analyzing the data has dropped dramatically, leading to a flood of investments in that area.
KerNEL is cognitive enhancement of the not-gimmicky variety. The concept is based on the work of Theodore Berger, a pioneering biomedical engineer who directs the Center for Neural Engineering at the University of Southern California, and is the startup's chief science officer.
This is a question of keeping humans front and center as we progress.
For over two decades, Berger has been working on building a neuroprosthetic to help people with dementia, strokes, concussions, brain injuries, and Alzheimer's disease, which afflicts one in nine adults over 65.
The implanted devices try to replicate the way brain cells communicate with one another. Let's say, for example, that you are having a conversation with your boss.
A healthy brain will convert that conversation from short term memory to long-term memory by firing off a set of electrical signals. The signals fire in a specific code that is unique to each person and is a bit like a software command.
Brain diseases throw these signaling codes off. Berger's software tries to assist the communication between brain cells by making an instantaneous prediction as to what the healthy code should be, and then firing off in that pattern.
In separate studies funded by the Defense Advanced Research Projects Agency over the last several years, Berger's chips were shown to improve recall functions in both rats and monkeys.
A year ago, Berger felt he had reached a ceiling in his research. He wanted to begin testing his devices with humans, and was thinking about commercial opportunities when he got a cold call from Johnson in October 2015.
He hadn't heard of Johnson; the Google search said he was a tech entrepreneur who had founded a payments processing company and invested in out-there science start-ups. The two met in Berger's office later that month. They talked for four hours, skipping lunch, and by end the day, Johnson said he would put up the funds for the two to start something together. "I don't know who, but somebody was looking over us," Berger said of the meeting.
For Johnson, the meeting was a culmination of a longtime obsession with intelligence and the brain.
I wanted to understand, what mental models people maintained - how did they define what to work on and why?
Shortly after he sold Braintree, he was already restless to start another company.
He spent six months calling everyone he knew who was doing "something audacious" - about 200 people in all. "I wanted to understand, what mental models people maintained - how did they define what to work on and why?" he says.
He then set up a $100 million fund that invests in science and technology start-ups that could "radically improve quality of life." The fund, which comes exclusively from his personal fortune, was called OS Fund, because he wanted support companies that were making changes at the so-called operating system level, he said.
Johnson's goal was to take projects from "crazy to viable" - including start-ups attempting to mine asteroids for precious metals and water, delivery drones for developing countries, and an artificial intelligence company building the world's largest human genetic database.
At the same time, he kept returning to intelligence, both artificial and real. As he saw it, artificial intelligence was booming - technology advances were moving at an accelerated pace; the pace of the human brain's evolution was sluggish by comparison. So he hired a team of neuroscientists, and tasked them with combing through all the relevant research, with the goal of forming a brain company. Eventually they settled on Berger.
Ten months later, the team is starting to sketch out prototypes of the device and is conducting tests with epilepsy patients in hospitals. They hope to start a clinical trial, but first they have to figure out how to make the device portable (Right now, patients who use it are hooked up to a computer).
Zoloth says one of the big risks of technologists funding science is that they fund their own priorities, which can be disconnected from the greater public good.
If brilliantly creative high school teachers in the inner city, for example, could fund science too, then perhaps the needs of the poor might be found more interesting.
Many people don't have enough resources to fulfill the brain potential they currently have, let alone enhance it. "Saying that if tech billionaires fund what they want may inadvertently fund science for the larger public, as a sort of leftover effect, is a problematic argument," she said.
"If brilliantly creative high school teachers in the inner city, for example, could fund science too, then perhaps the needs of the poor might be found more interesting."
Johnson says he is acutely aware of those concerns. He recognizes that the notion of people walking around with chips implanted in their heads to make them smarter seems far-fetched, to put it mildly. He says the goal is to build a product that is widely affordable, but acknowledges there are challenges.
He points that many scientific discoveries - even the printing press itself - started out for a privileged group but ended up providing massive benefits to humanity.
The primary benefits of KerNEL, he says, will be for the sick, for the millions of people who have lost their memories due to brain disorders. Even a small improvement in memory - a person with dementia might be able to remember the location of the bathroom in their home, for example - can help people maintain their dignity and enjoy a greater quality of life.
And in an age of AI, he insists that boosting the capacity of our brains is itself an urgent public concern. "Whatever endeavor we imagine - flying cars, go to Mars - it all fits downstream from our intelligence," he says. "It is the most powerful resource in existence. It is the master tool."