Chances are, if you’ve used the internet much in the past decade, you’ve heard of algorithms — those lines of computer code that curate and control data, like a recipe lists the ingredients of a meal and tells us how to put it all together.
On our phones, tablets and other devices in our digital world, algorithms hail our Uber rides, decide who we should befriend on social media, and choose shows, movies and products for us to consume. They can even help companies decide what candidates to hire and help universities diversify their student bodies. One day, automakers hope, complex algorithms will drive cars more safely than humans do.
On paper, algorithms sound like the pinnacle of efficiency, but as they’ve become more ubiquitous, they’ve also come under scrutiny for taking important decisions out of the hands of humans.
Christine Weltner demonstrates cursive writing to her third-grade class on March 1, 2017, in Queens, New York. A controversial algorithm employed in New York City schools grades teachers based on their effectiveness. | Mary Altaffer, Associated Press
In her 2016 book, “Weapons of Math Destruction,” which was longlisted for the National Book Award, data scientist Cathy O’Neil makes a compelling argument that humans’ trust in algorithms is misplaced, or at least premature. She points to the algorithm employed in New York City schools’ controversial value-added model of grading teachers based on their effectiveness, dictating job security.
Other examples O’Neil singles out include algorithms employed to calculate likely recidivism rates of convicted inmates, which ProPublica found falsely attributed race as a factor in whether an inmate re-offends, with African-Americans more likely and whites less likely to be repeat offenders.
“The consequences are serious because the models are used in each stage of the criminal process by local judges and contribute to determining everything from bond amounts to sentencing,” Quartz summarized.
Looking to the future of how algorithms stand to impact everyday life, Pew Research Center released a report in February cataloging the hopes and concerns of about 1,300 technology and digital data scholars and experts, describing what some of them call “The Age of Algorithms.”
Here are some of the main takeaways from Pew’s report:
Checks and balances
Although algorithms are used in global finance and other ways that impact society, there is no legal recourse or official authority to hold a company responsible for the actions of its algorithms. That’s partially because algorithms are often developed and implemented in secret to avoid hacking and reverse engineering. But it’s also because the public and government are only beginning to realize the dangers and complexity of algorithms.
As Pew pointed out, the Obama administration outlined some of those dangers in a report and accompanying strategic plan last fall, which called for extensive research into the reach and implications of algorithms.
Understanding algorithms and their impact on human life goes far beyond basic digital literacy, some experts said.
“We should question how our data and decisions are made as well. What is the supply chain for that information? Is there clear stewardship and an audit trail?” Susan Etlinger, an industry analyst at Altimeter Group, told Pew. “The power to create and change reality will reside in technology that only a few truly understand.”
Many experts included in Pew’s report called for authorities to go further, establishing a regulatory framework — what researcher Andrew Tutt called “The FDA for Algorithms” — that could harness the usefulness of algorithms while safeguarding against the biases and vulnerabilities they create.
“Algorithmic regulation will require federal uniformity, expert judgment, political independence and pre-market review to prevent — without stifling innovation — the introduction of unacceptably dangerous algorithms into the market,” Tutt said.
Until such regulations are implemented, there’s little accountability for an algorithm’s flaws, and little recourse for the people who may suffer as a result.
While much of the report focused on the dangers and downsides of algorithms, many experts predict algorithms will improve life in endless ways, from medicine to transportation to governance.
Stephen Downes, from the National Research Council of Canada, was Pew’s lead proponent of the potential improvements algorithms may bring.
“Health care is a significant and growing expense not because people are becoming less healthy (in fact, society-wide, the opposite is true) but because of the significant overhead required to support increasingly complex systems,” Downes told Pew. “New technologies will enable health providers to shift a significant percentage of that load to the individual, who will … coordinate and manage their own care, and create less of a burden on the system.”
Downes also sees algorithms making American politics more responsive to voters by allowing technology to create electorate boundaries based on census data.
“One of the most persistent political problems in the United States is the gerrymandering of political boundaries to benefit incumbents,” Downes said. “Electoral divisions created by an algorithm to a large degree eliminate gerrymandering.”
Large scale manipulation
As fast as artificial intelligence and algorithms continue to advance, some experts worry humans will increasingly be left to the mercy of whomever controls the technology.
“My biggest fear is that … it will be simply too convenient for people to follow the advice of an algorithm (or, too difficult to go beyond such advice), turning these algorithms into self-fulfilling prophecies, and users into zombies who exclusively consume easy-to-consume items,” said Bart Knijnenburg, a Clemson University human-centered computing professor.
Other experts, whose comments Pew included anonymously, questioned what happens to humanity and moral values when society increasingly “lets the robots decide.”
Facebook's news algorithm tailors updates and content to each user, which could keep users from receiving information or news from sources that challenge their worldview. | Deposit Photos
An example could be the recent outcry over Facebook's news algorithm, which enhances the so-called “filter bubble” of information. The social media giant's news algorithm tailors updates and content to each user, which could keep users from receiving information or news from sources that challenge their worldview.
“The fact that the internet can, through algorithms, be used to almost read our minds means [that] those who have access to the algorithms and their databases have a vast opportunity to manipulate large population groups,” one anonymous expert said.
“By utilitarian metrics, algorithmic decision-making has no downside; the fact that it results in perpetual injustices toward the very minority classes it creates will be ignored,” another wrote. “The Common Good has become a discredited, obsolete relic of The Past.”
Human vs. machine-driven economy
Algorithms have already led to automation for many jobs, including mass transit conductors and factory jobs, but they stand to take over many other careers as well — including writers and even physicians.
“With the rise of the algorithm — humans will be replaced by machines/computers for many jobs/tasks,” one anonymous CEO told Pew. “What will then be the fate of Man?”
As this trend continues, experts told Pew, the world may have to pump the brakes on tech innovation in the economy and actively choose which jobs will be populated with human workers, and which will become the domain of robots.
“Every time you design a human system optimized for efficiency or profitability you dehumanize the workforce,” an anonymous futurist told Pew. “When you remove the humanity from a system where people are included, they become victims.”
An autonomous vehicle is tested by an engineer on a street through an industrial park in Boston on Jan. 10, 2017. Researchers are constantly creating better algorithms to guide autonomous vehicles. | Steven Senne, Associated Press
Others say an economic chasm will form between those who can fully automate their businesses and those who rely on a human workforce.
“The massive boosts in productivity due to automation will increase the disparity between workers and owners of capital,” one user wrote.
Those hardest hit will be the most vulnerable among Americans — the poor and the undereducated who depend on non-tech jobs to survive.
“Modern Western society is built on a societal model whereby capital is exchanged for labor to provide economic growth,” one respondent said. “If labor is no longer part of that exchange, the ramifications will be immense.”
Yet, some experts speculate that society’s own sense of survival will promote more of a balance between technology and humans before the economy becomes the stuff of science fiction. Humankind, after all, is still in charge of technological innovation for the time being, with the ability to correct course as the digital landscape of algorithms changes.
“We are finally reaching a state of symbiosis or partnership with technology. The algorithms are not in control; people create and adjust them,” Microsoft principal researcher Jonathan Grudin told Pew. “I’m optimistic that a general trend toward positive outcomes will prevail.”