The digital transition was always going to be a messy one—look at the antitrust ﬁghts that followed the telephone during the analog era. "System Error," by Rob Reich, Mehran Sahami and Jeremy Weinstein.
September 23, 2021
Facebook’s Mark Zuckerberg testifies before the House Financial Services Committee in 2019. Photo: Chip Somodevilla/Getty Images
HOW TO FIX THE INTERNET? It’s all so new—the internet, and digital technology in general—that you sometimes wonder how something so promising could have gone wrong so quickly. But there is no precedent for the world we’re creating. The internet changes everything, as its fanboys used to say, back before the internet actually did change everything.
That was 20-odd years ago, when mass media were at their apogee. Television, newspapers, magazines, the music industry—all were bigger and more flush than they’d ever been. The dot-com bust left the internet economy in a shambles, convincing doubters that the whole thing was a fad. In reality, the bust simply cleared out a raft of unsustainable enterprises—Pets.com, anyone?—and set the stage for what was to come.
Today, thanks to broadband, music has been reinvented as a streaming service and television is fast doing the same. Most newspapers have been abandoned by readers and advertisers alike. Alphabet, Amazon, Apple, Facebook and Microsoft dominate Wall Street. The internet, with some help from the pandemic, has fundamentally changed the way we shop, work, get our news, get our entertainment and engage with one another. We now have a supercomputer in our pocket—along with surveillance systems that supposedly tell advertisers what we want, software that can pick out a face in a crowd, hiring algorithms that can perpetuate human biases at scale, and black-box recommendation engines that can send you to prison but can’t say why.
SYSTEM ERROR: Where Big Tech Went Wrong and How We Can Reboot by Rob Reich, Mehran Sahami and Jeremy M. Weinstein Harper, 352 pages, $27.99
So it’s heartening to see three professors— Rob Reich, a philosopher; Mehran Sahami, a computer scientist; and Jeremy Weinstein, a policy wonk—try to figure it out in System Error: Where Big Tech Went Wrong and How We Can Reboot. Early on, the authors—all three from Stanford, the wellspring of Silicon Valley—introduce us to techie concepts like “optimization,” the drive to achieve maximum efficiency with little regard to what is being optimized or why.
They tell us of Joshua Browder, a former Stanford student who created a chatbot to help people like himself get out of paying parking tickets. Mr. Browder had racked up a lot of them, and though he admitted that he “probably deserved them,” he couldn’t afford to pay up. His chatbot built up a database of reasons parking tickets get thrown out, then provided users with “the optimal grievance procedure” to appeal. This caught the attention of the folks at Andreessen Horowitz, one of the Valley’s premier venture-capital firms, which helped Mr. Browder turn his self-serving toy into a business. Mr. Browder is now the CEO of DoNotPay, a company that’s raised nearly $28 million in funding and uses artificial intelligence to offer people ways out of all sorts of scrapes.
Then there’s another ex-Stanford student, Aaron Swartz, a co-founder of Reddit who cared deeply—maybe too deeply—about what technology is used for. One of his causes was open access to scholarly journals. In 2011 he was arrested and charged with using the MIT computer network to download millions of papers from JSTOR, the fee-based online academic database, apparently intending to release them for free. JSTOR declined to press charges, but federal prosecutors piled them on anyway. Swartz, facing up to 50 years in prison, hanged himself at the age of 26.
While Swartz has largely been forgotten, Mr. Browder has joined the pantheon of Silicon Valley entrepreneurs—even as the Valley itself has become less entrepreneurial and more about being home to the world’s most-valued corporations. But if tales like these illustrate the shift in ethos from Steve Jobs’s “think different” to Mark Zuckerberg’s “move fast and break things” mantra, they don’t tell us how to reboot.
That’s a far tougher question, one that requires a ground-up rethink of matters ranging from Section 230 of the 1996 Communications Decency Act—which absolves internet publishers of responsibility for almost anything their users post—to the focus on maximizing shareholder value above all else. If you run a social-media site in such a world, your goal is to increase ad revenue by keeping users engaged at all costs. Which is what turned Facebook, for example, into a miasma of misinformation, disinformation, hate speech and anger that bizarrely coexists with a loving panoply of cat vids and baby pics.
Likewise, our current conception of antitrust, with its focus on protecting consumers from price gouging, leaves us at a loss to deal with giant corporations that make their platforms available to users for free while squelching or buying up potential competitors. And when American privacy laws permit those corporations to sell their users’ data to the highest bidder, consumers are left defenseless from the real danger, which is digital surveillance and data harvesting.
Surveillance, competition, monopoly, social media, online bullying, artificial intelligence, the optimization mindset—there are no easy answers for any of these, and most of the fixes that System Error offers are so high-minded and general that you have to wonder whether they’ll work. Are there answers at all? Most politicians are out of their depth on technology issues, and most technologists seem too enamored of the tech and its material benefits to have the requisite perspective. Not that all techies are as self-involved as the authors suggest: Employees at Google have spoken up against the company’s failure to stem homophobic harassment at its YouTube subsidiary; data researchers at Facebook have tried in vain to address their news feed’s tendency to promote toxic content and the negative effect of Instagram on teen girls.
This book’s contribution, and it’s an important one, is to spell out what needs to be fixed, and to provide two important reminders. First, the transition to digital in a democratic society was going to be messy no matter what—look at the antitrust fights that accompanied the telephone in the analog era. And second, solutions are going to be less about achieving utopia than about setting parameters to make sure certain things don’t happen. We live at best in a good-enough world. ◼︎