Silicon Valley is going to face real competition in several key technical areas in years to come — a key take-away from a lively talk by former New York Times science and technology writer John Markoff, who spoke before more than 100 people at the Portola Valley Community Hall on Oct. 17.
In his hour-long talk, Mr. Markoff discussed big changes ahead, but not before a short tour of his 28 years of writing stories about Silicon Valley. He noted encounters with notorious hackers, and writing breaking stories — on the debut of the World Wide Web, for example, and a proposed government encryption backdoor known as the clipper chip.
The big changes included an emerging global leader in artificial intelligence research that is not the United States, the approaching limits of Moore’s Law, and a significant shift of the Bay Area’s center of venture capital investment in high technology.
That center is now San Francisco, Mr. Markoff said, citing an analysis by urban studies theorist Richard Florida.
In an Oct. 3 article in the online publication CityLab about a rumored relocation of startups and talent to less expensive locations, Mr. Florida includes Oakland and Hayward with San Francisco as representing 34 percent ($23.4 billion) of U.S. VC high-tech investment in 2016. He assigns just 9.8 percent ($6.7 billion) to the metro area of San Jose, Sunnyvale and Santa Clara.
Looking back and using VC investment as a gauge, Mr. Florida determined that the center of Silicon Valley in 1984 was the city of Santa Clara, Mr. Markoff said. Jumping ahead another decade or so, Mr. Markoff added a historical note: Silicon Valley had begun to slow down in 1996, and by 2006, Europe was outdoing Silicon Valley in innovation. But the 2007 debut of the Apple iPhone turned things around, he said.
“The question is, ‘What’s going to happen next time?'” he said.
Moore’s Law
The end may be near for the reliable physics of the manufacture of semiconductors.
Moore’s Law, a mid-1960s prediction by Intel Corp. co-founder, philanthropist and Woodside resident Gordon Moore, posited that continuous technological advances would allow the doubling of the number of transistors on an integrated circuit every two years. This insight has proven accurate and has enabled decades of faster, smaller and cheaper computers.
But efficiency limits are within sight, maybe not forever but for now, Mr. Markoff said. The rate of decline in the cost of transistors began to slow around 2012, he said. In 2015, he said, Intel missed its target for increasing the transistor density on a chip (per Moore’s Law). The two-year cycle is now closer to two and a half years, Intel CEO Brian Krzanich said in a July 2015 online story in The Verge.
And there’s “dark silicon,” the industry’s term for integrated circuits in which the transistors are packed so densely that activating all of them at once will melt the device, Mr. Markoff said.
Not necessarily bad news, he noted, recalling an exchange with a Stanford University student who was “over the moon” about approaching this limit. “Now it’s our turn,” Mr. Markoff recalled the student saying. “The free ride is over. Now it’s going to have to be human creativity if we’re going to go any further.”
China and AI
If all this isn’t enough, Silicon Valley is facing a serious challenge in the research and development of artificial intelligence. The Chinese government has declared AI a nationwide priority, Mr. Markoff said. It’s an “entrepreneurial frenzy” over there, he said, with 4 million startups in 2016, 1.7 million of them in high technology. “It felt like Silicon Valley had infected China,” he said of his visit there.
Silicon Valley tends to underestimate the importance of government funding, he said, particularly during the era leading up to the 1980s, when key funding came from the Defense Advanced Research Projects Agency.
“All the seed technologies that have created these platforms we live on (today) were due to DARPA funding,” Mr. Markoff said. “So just at the juncture where we’re about to have a real competitor, it seems like our government is abandoning the field. We won’t find out what the consequence is for probably a decade, but it’s an interesting time.”
China has a vested interest in AI and robotics, with the working-age workforce shrinking “dramatically quickly,” Mr. Markoff said. He recalled a conversation with Nobel Prize winning economist Daniel Kahneman who, when Mr. Markoff expressed concern about robots and their threats to jobs in China, replied: “You don’t get it. In China, they’ll be lucky if robots arrive just in time.”
The root of the problem, Mr. Kahneman explained, is China’s one-child policy — introduced in 1979, relaxed in 2013 and discontinued in 2015.
But while the working-age workforce is shrinking around the world, the more important number, Mr. Markoff said, is the dependency ratio — the number of people able to give care compared to the number of people who will need care.
At least one roboticist commented to Mr. Markoff that self-driving cars will be the first elder-care robots. He disputes that. “Tell me when a robot will be able to safely give an aging human a shower,” he said. “That’s the benchmark I’m interested in and I think that’s a harder challenge for society.”
AI’s threats
In the Silicon Valley of the 1980s and 1990s, if there was an ethos, it’s expression might have been captured by the title of the 2004 book: “The Inmates are Running the Asylum,” by Alan Cooper, a leader in advancing the understanding and importance of design in crafting interactions between humans and computers.
People in Silicon Valley didn’t really understand design back then, Mr. Markoff said in an interview. There was no science of design, no theory of design. “The technology was ahead of people’s understanding of how to use it,” he said.
The design of the computer mouse was a case in point. Douglas Englebart, an Atherton resident who died in 2013, invented the mouse in the mid-1960s and was an advocate of lots of buttons, Mr. Markoff said. Steve Jobs, designer of the Apple Macintosh computer in the 1980s, said he wanted just one “because then you couldn’t push the wrong button,” Mr. Markoff said.
Is a chaotic debut also ahead for artificial intelligence? It’s less likely, Mr. Markoff said. There are a group of people, among them Eric Horvitz, a technical fellow at Microsoft Corp., who are looking ahead and focused on AI as augmentation to human capacity rather than replacement. The idea of designing with ethical values is gaining traction, Mr. Markoff said, adding: “That a good sign, I think.”
Not that there’s nothing to worry about. Mr. Markoff noted recent controversial research by Stanford scientists who designed an algorithm that could distinguish between gay and straight men based on facial images. Surveillance and the limits of privacy will be issues, he said. It won’t be long, he added, before computers can imitate an individual’s speaking voice, a handy tool to perpetrate fraud.
“Those kinds of things are going to push up against the ethical border really quickly,” he said.
AI will present interesting questions if the government decides to regulate it, he said. Will machines masquerading as humans be detectable? Will AI be able to distinguish between good and bad actors?
Asked whether these kinds of questions kept him up at night, Mr. Markoff said they did not. What does worry him is the current national political situation, the potential for nuclear conflict and “whether we’ll walk into wars with a stupid purpose,” he said.
‘The end of an era’: After 36 years in Palo Alto, Il Fornaio prepares to closeJanuary 13, 2026 2:05 pm
San Mateo County taps community affairs director as interim head of Parks DepartmentJanuary 12, 2026 4:54 pm