Artificial intelligence (AI) will shape the destiny of humanity, but first humanity has the opportunity to shape AI. At times, anxiety about the former causes us to overlook the latter. We forget that artificial intelligence is going to serve the goals with which we’re now programming it. Its implementation will follow the standards we now have the opportunity to set.
This leaves citizens — and those representing us in government — with an urgent responsibility. We’ll need robust strategies that ensure the U.S. is positioned to pioneer these new technologies, reaping their profound benefits while establishing precedents for their responsible use. We’ll also need public policy frameworks that prepare us for the problematic repercussions accompanying their introduction.
We’ve been slow to commence constructive dialogue on these matters, but one year out from an historic election, Democratic candidates are at least beginning — finally — to engage the public on AI. Though the conversation is still scant and underdeveloped, at this critical juncture, it’s essential to pay close attention to the emergent political discourse, to who’s saying what on AI, and to place it into context.
An overdue national strategy
There’s a reason Russian President Vladimir PutinVladimir Vladimirovich PutinUkraine, Russia agree to restart peace process Trump, Russian foreign minister to meet Tuesday Impeachment, Ukraine, Syria and warheads color Washington visit by top Russian diplomat MORE did not raise many eyebrows when he proclaimed that the nation leading the way in AI “will become the ruler of the world.” And in China, recognition of AI’s potential to revolutionize its economy, systems of governance, and military is leading the superpower to pour billions into AI-related companies.
Meanwhile, the U.S. is at risk of falling behind, potentially depriving our economy of 15 percent growth in GDP over the next decade, compromising our national security, and forgoing our chance to be the global trendsetter in harnessing AI for the good of society. Once behind in the race, playing catch-up with an increasingly intelligent entity will likely prove impossible.
There is, therefore, a dire need for a comprehensive national AI strategy that prioritizes investment and implementation of the technology both in the public and private sectors. This necessity hasn’t escaped even the current U.S. administration, which has otherwise cut funding for basic science and technology research. Building on initiatives from Obama’s presidency, President TrumpDonald John TrumpLawmakers release defense bill with parental leave-for-Space-Force deal House Democrats expected to unveil articles of impeachment Tuesday Houston police chief excoriates McConnell, Cornyn and Cruz on gun violence MORE introduced the American AI Initiative earlier this year, a modest plan for funding AI R&D while also implementing U.S.-led international AI standards.
Though applauded by experts as a step in the right direction, it has also been criticized as somewhat vague and noncommittal.
So, what do Trump’s Democratic rivals have to offer as an alternative?
Unfortunately, while most have acknowledged the need for a broad national AI plan, few have offered even hints of what such a strategy might look like.
South Bend, Ind. Mayor Pete ButtigiegPeter (Pete) Paul ButtigiegGabbard says she won’t participate in next debate even if she qualifies McKinsey allowing Buttigieg to disclose past clients Saagar Enjeti: Elizabeth Warren reveals grim future under her presidency MORE has argued on occasion that instead of relying solely upon the private sector to drive innovation in the field, the government must dramatically increase spending on AI research and development.
But it is only former Maryland Congressman John DelaneyJohn Kevin DelaneyDelaney to DNC: Open second debate stage for candidates who qualified for past events Krystal Ball: What Harris’s exit means for the other 2020 candidates 2020 Democrats thank Harris for friendship, candidacy after senator drops out MORE who’s selling voters a fully-fledged national AI strategy — and voters should take note. The founder and former co-chair of the House Artificial Intelligence Caucus, Delaney has trumpeted the dawn of AI as a historic opportunity capable of yielding unprecedented benefits to society. His proposed plan would dramatically increase government research funding, woo high-tech professionals into government positions, welcome talented immigrants and set up an interagency to consider how to manage the drawbacks of adoption.
Compared with prevailing rhetoric, Delaney’s cautious optimism is as commendable as the thoroughness of his plan.
In the international AI race, the U.S. must appreciate its opportunity to responsibly lead the world in innovation. Voters would be wise to ask their candidates for plans detailing how the U.S. can set an example through initiatives that establish AI as a force for social good — from ensuring food security and tackling climate change, to precision medicine and inclusive economic growth.
Talking privacy and bias
What the public stands to gain from AI, then, will hopefully captivate voters as much as the more highly publicized hazards of the new technology. Even so, a candidate’s plans for guarding against the negative repercussions of AI should be central to their platform.
As public awareness grows, privacy concerns over facial recognition surveillance systems are mounting. Biometric software may help police track down burglars, or even preempt a terrorist attack, but tech that watches us can engender fear and coerce us towards self-censorship. Voters are equally alarmed by the proneness to bias of the highly advanced algorithms currently helping public and private institutions decide how they hire, offer loans, advertise, and even pass down criminal sentences. A ProPublica investigation, for example, revealed implicit biases in courtroom algorithms that resulted in harsher sentences for people of color.
Almost all the Democratic presidential hopefuls have addressed these growing anxieties, some providing more clues than others as to how they would tackle the issues if elected.
Sen. Cory BookerCory Anthony BookerRosario Dawson joins Booker for ‘Lead with Love’ tour Sunday shows — Nadler: A jury would convict Trump in ‘three minutes flat’ Booker on Harris dropping out: ‘Iowa voters should have the right to choose’ MORE (D-N.J.), and Sen. Elizabeth WarrenElizabeth Ann WarrenGabbard says she won’t participate in next debate even if she qualifies On The Money: White House, Dems edge closer to trade deal | GOP worries about Trump concessions | DOJ argues Congress can’t sue Trump on emoluments | Former Fed chief Volcker dies UN International Anticorruption Day highlights democracy as a human right MORE (D-Mass.) both signed letters insisting that the FBI and all federal agencies regulating job discrimination and economic opportunity address questions pertaining to algorithmic bias and facial recognition systems. In April, Booker went one step further when he joined other U.S. lawmakers in introducing the Algorithmic Accountability Act, aimed at forcing the Federal Trade Commission to help audit potentially biased machine learning algorithms used by large companies.
The most extreme — and perhaps counterproductive — stance so far was assumed by Sen. Bernie SandersBernie SandersLawmakers release defense bill with parental leave-for-Space-Force deal Gabbard says she won’t participate in next debate even if she qualifies Overnight Health Care — Presented by That’s Medicaid — Deal on surprise medical bills faces obstacles | House GOP unveils rival drug pricing measure ahead of Pelosi vote | Justices to hear case over billions in ObamaCare payments MORE (I-Vt.), who called for a total ban on police use of facial recognition software. Outlining a broader criminal justice reform plan, he also recommended a moratorium on courtrooms’ use of algorithms predicting recidivism.
The next occupant of the White House should implement protocols for multi-level audits of machine learning models rather than banning them outright. Internal reviews of newly developed models could perhaps be supplemented by third-party oversight from consortiums consisting of companies, universities, civil rights groups, and an AI government advisory council.
Dialoguing on job displacement
While AI adoption will eventually usher in debates on everything from national defense to fair competition practices, there is no issue that currently galvanizes the public like automation’s effect on jobs.
Estimates putting job loss at as high as 14 percent by 2030 have evoked widespread angst. Meanwhile, a recent survey suggests that Democratic-leaning voters are nearly twice as likely to expect the government to step in if workers are displaced by robots.
The future of work has therefore sparked spirited debate and elicited a wide range of diverse proposals. Perhaps no one has drawn more attention to the issue than Andrew YangAndrew YangGabbard says she won’t participate in next debate even if she qualifies Emanuel jokes: ‘I’m a new, mellow Rahm’ Yang expands campaign with senior hires for digital operations MORE, who has painted a bleak picture, warning of the very unravelling of our society while arguing that the remedy can be found in a Universal Basic Income.
Striking a somewhat less alarming tone, Sanders promotes his federal jobs guarantee as the most viable solution to offset any job loss effected by automation. Meanwhile, tech trustbuster Warren has actually downplayed the effect of automation, naming bad trade policy as the primary culprit behind job loss.
The most salient point may have been tweeted by Delaney, who called the idea that automation eliminates all jobs “a complete fantasy,” stating that “Innovation always displaces AND creates jobs.” Delaney, like Buttigieg and Joe BidenJoe BidenGabbard says she won’t participate in next debate even if she qualifies House Democrats expected to unveil articles of impeachment Tuesday FBI head rejects claims of Ukrainian 2016 interference MORE, have argued for retraining programs that equip people with the skills necessary for information era redeployment. A far less flashy argument, but it’s one that shouldn’t be overlooked.
What should also not be overlooked is just how important it is that conversations related to public policy on AI are finally being brought center stage.
In a world where populist arguments exploiting broad fears have recently trumped more thoughtful, solution-oriented thinking, voters should be careful to familiarize themselves with the current conversation and to help ensure that dialogue remains constructive.
Artificial intelligence does pose certain perils, but it also holds tremendous promise — that is, if our leaders act quickly and practically to set the right precedents.