What Might the Future Hold?

Print view

At Stanford's Future Fest, a group of engineers from industry and academia discuss artificial intelligence, the "revolution" in biology and the balance between risk and regulation.

Dean Persis Drell and Astro Teller. Photo by Matt Beardsley Creative

Stanford Engineering Dean Persis Drell, left, with Google's Astro Teller, at the Stanford Technology Venture Program's Future Fest | Photo by Matt Beardsley Creative

Futurology confounds optimists and pessimists alike. Where are the personal jetpacks and flying cars predicted in the 1950s? Dystopian pronouncements are equally error prone. Remember the jeremiads about peak oil and $150-a-barrel crude? Then came the fracking boom.

Even people who help shape the future can botch near-term predictions. In 1995, for example, 3Com founder and Ethernet inventor Robert Metcalfe wrote, “The Internet will soon go spectacularly supernova and in 1996 collapse.” Four years later Metcalfe literally ate his words by stuffing a copy of his errant column into a liquid-filled blender at the Sixth International WWW Conference and gulping down a humble pie puree.

Still the future beckons, and prognosticators stand on safer ground when they extrapolate probabilities from emerging trends rather than predict specific developments at particular times. That was the ethos of Future Fest, a conference held under the auspices of the Stanford Technology Ventures Program.

In an effort to tease out future trends, Stanford Engineering Dean Persis Drell engaged in conversation with three tech superstars: Google’s “Captain of Moonshots” Astro Teller; Stanford bioengineer Christina Smolke; and Silicon Valley venture capitalist Steve Jurvetson, a partner at DFJ.

From amongst the wisps of possibility, some probabilities emerged, including:

The likelihood that rapid change will dictate shorter lifespans for any given technology …

         

It’s not just that technology is evolving rapidly, observed Jurvetson: The rate of that change is accelerating.

“The corollary is that the lifespan of any new technology or product will become increasingly ephemeral,” said Jurvetson.

That applies as well to the hard science that underlies all technology, added Smolke:

“What is cutting edge now will be obsolete in 10 years,” she said. “[For example,] our current models of genetics were different 25 years ago, and were taught differently.”

The rise of software and the emergence of artificial intelligence …

    

Citing the ubiquity of robotics in various fields, Teller observed there is an emerging “commonality” to technology.

Learning about cutting-edge technologies at the Stanford Technology Venture Program's Future Fest . Photo by Matt Beardsley Creative“It’s not traditional robotics but it’s still robotics,” he said. “Self-driving cars are nontraditional robots. Package-delivery drones are essentially robots. So are airborne wind turbines [used on stratospheric balloons Google is developing]. We’re bringing different techniques from machine learning, from AI [artificial intelligence], from mechanical engineering to embed intelligence into products.”

Jurvetson noted that broad technological fields once considered distinct and unattractive are rapidly modernizing: “Rockets, cars, satellites, robots—they used to be industrial, and are now software-centric.”

 

Coupled with the dawning of a new, golden age of biotechnology …

       

Despite the accelerating integration of electronic technologies, dramatic breakthroughs in biotechnology are creating the foundation for the next great wave of scientific advances. Among the developments supporting this trend, said Smolke, are cloud biology laboratories—virtual labs that allow researchers to outsource and systematize their experiments easily and rapidly, providing a level of automation and scalability new to biological experimentation.

“A revolution is happening in biology with genetics and genomics,” said Smolke. “Traditionally, biology has been a niche; many engineers don’t take biology classes. We need to change that.” A broad biological curriculum, continued Smolke, will prepare engineers for the “next realm of technological advances—the melding of biology with computer science, information science and mechanics.”

Progress is so rapid in biotech that some popular fictions are on the cusp of becoming realities, Smolke noted, citing de-extinction—the re-creation of long-vanished beasts and plants—as an example.

“We’re not there yet, but we’re inching closer with different creatures,” she said.

A recognition that learning to learn may be more important than learning specific things …

       

Given that much of the knowledge that an aspiring engineer or researcher acquires today will be obsolete in a decade, said Teller, the single most critical skill is learning resiliency.

Learning about cutting-edge technologies at the Stanford Technology Venture Program's Future Fest . Photo by Matt Beardsley Creative“The ability to learn things quickly, the skill to reconstruct first principles after you forget 90 percent of them later, those skills will matter more than the actual knowledge you’ll be taking in,” he said.

Jurvetson concurred: “Process learning will be more important than product learning,” he said.

And as technology advances, the stereotype of the lone researcher achieving dramatic breakthroughs will seem increasingly antiquated, said Smolke; the future belongs to the collaborators.

“We teach students to work in interdisciplinary teams,” she said. “They’re going to have to be able to leverage each other’s expertise, and be good at hand-offs and communication.”

Researchers and regulators will be locked in an uneasy symbiosis …

           

Jurvetson observed that regulators ought to start from the principle of the Hippocratic Oath: “First do no harm.” That said, he continued, government can play a large and productive role in helping technologies in the critical start-up phases, e.g., NASA and space exploration, and loans to automotive and energy storage company Tesla at critical junctures. Government grants to research universities will also remain central to technological progress.

Teller said Google constantly interacts with government agencies and considers their role generally salutary—or at least tolerable. The best technique for dealing with regulators, he said, is to approach them well in advance with general goals rather than presenting them with a finished product.

“We say, ‘This is what we’re working on—what do you think?’” said Teller. “They can then partner with us in the creative process.”

Future risks are real and must be addressed …

       

The panelists acknowledged that concerns over the possible downsides of future tech are not misplaced. Compounding advances ultimately could lead to powerful technologies “that could be trivial for an undergraduate to develop [and that could] kill much of humanity for almost no money,” Jurvetson observed.

Human cultural evolution, Jurvetson continued, is “pretty static. It’s glacial compared to technological advancement. Technology is synonymous with progress, [and] I think it’s a net positive, but there are some big implications [for social disruption]. I worry about the accelerating gap between rich and poor. As business becomes more info-centric, we won’t see jobs in agriculture. Driving—which accounts for 20 percent of paid jobs today—will be going away in the near term. What can an individual or technology [sector] do? I don’t know, unless an entrepreneur tries to find a solution, a way to provide for all.”

Smolke said that technologists are responsible for the broad social impacts of their work, “so it partly comes down to our research choices. There are dual-use aspects of many technologies. You can’t predict all the ways a technology will evolve or how people will use it. But you can do some near-term predictions, and you can decide if the overall risk is acceptable. You can work with regulators to ensure policies that will mitigate dangers, and you can also train students to innovate positively.”

Still, emphasized Teller, researchers mustn’t be expected to control outcomes: “That’s the job of the public sector. Our job has to be transparency and education. Instead of presenting technologies as poster children for our fears, we should be educating and moving forward.”

But the singularity is not necessarily near …

   

While the idea of a “singularity”—the abrupt emergence of a powerful and sentient AI with its own agenda—has been hypothesized by computer scientist and futurist Ray Kurzweil, Teller maintained that most scientists are highly skeptical of what some call “transhumanism.”

The panelists downplayed the imminence of super-intelligent and autonomous AI or similarly dramatic developments, such as transferring a human consciousness into a cybernetic system or “cut-and-paste” interfaces that would allow people to download martial arts skills, foreign language fluency or other knowledge directly into their brains.

“You’re not wasting your time going to school, and your grandkids will not be wasting their time going to school,” Teller said.

View a video of the full 48-minute panel discussion, “Forecasting the Future of Technology,” on the Stanford Entrepreneurship Corner website.

Last modified Wed, 10 Feb, 2016 at 14:17