Skip to content
Search

Latest Stories

Top Stories

Avoiding disaster by mandating AI testing

Avoiding disaster by mandating AI testing
Getty Images

Kevin Frazier will join the Crump College of Law at St. Thomas University as an Assistant Professor starting this Fall. He currently is a clerk on the Montana Supreme Court.

Bad weather rarely causes a plane to crash — but the low probability of such a crash isn’t because nature lacks the power to send a plane woefully off course. In fact, as recently as 2009, a thunderstorm caused a crash resulting in 228 deaths.

Instead, two main factors explain why bad weather no longer poses an imminent threat to your longevity: first, we’ve improved our ability to detect storms. And, second and most importantly, we’ve acknowledged that the risks of flying through such storms isn’t worth it. The upshot is that when you don’t know where you’re going and if your plane can get you there, you should either stop or, if possible, postpone the trip until the path is in sight and the plane is flight worthy.

The leaders of AI look a lot like pilots flying through a thunderstorm — they can’t see where they’re headed and they’re unsure of the adequacy of their planes. Before a crash, we need to steer AI development out of the storm and onto a course where everyone, including the general public, can safely and clearly track its progress.

Despite everyone from Sam Altman, the CEO of OpenAI, to Rishi Sunak, the Prime Minister of the UK, acknowledging the existential risks posed by AI, some AI optimists are ignoring the warning lights and pushing for continued development. Take Reid Hoffman for example. Hoffman, the co-founder of LinkedIn, has been "engaged in an aggressive thought-leadership regimen to extol the virtues of A.I” in recent months in an attempt to push back against those raising redflags, according to The New York Times.

Hoffman and others are engaging in AI both-side-ism, arguing that though AI development may cause some harm, it will also create societally beneficial outcomes.The problem is that such an approach doesn’t weigh the magnitude of those goods and evils. And, according to individuals as tech savvy as Prime Minister Sunak, those evils may be quite severe. In other words, the good and bad of AI is not an apples-to-apples comparison -- it’s more akin to an apples to obliterated oranges situation (the latter referring to the catastrophic outcomes AI may lead to).

No one doubts that AI development in “clear skies” could bring about tremendous good.For instance, it’s delightful to think of a world in which AI replaces dangerous jobs and generates sufficient wealth to fund a universal basic income.The reality is that storm clouds have already gathered.The path to any sort of AI utopia is not only unclear but, more likely, unavailable.

Rather than keep AI development in the air during such conditions, we need to issue a sort of ground stop and test how well different AI tools can navigate the chaotic political, cultural, and economic conditions that define the modern era. This isn’t a call for a moratorium on AI development -- that’s already been called for (and ignored). Rather, it’s a call for test flights.

“Model evaluation” is the AI equivalent of such test flights. The good news is researchers such as Toby Shevlane and others have outlined specific ways for AI developers to use such evaluations to identify dangerous capabilities and measure the probability of AI tools to cause harm in application. Shevlane calls on AI developers to run these "test flights", to share their results with external researchers, and to have those results reviewed by an independent, external auditor to assess the safety of deploying an AI tool.

Test flights allow a handful of risk-loving people to try potentially dangerous technology in a controlled setting. Consider that back in 2010 one of Boeing's test flights of its 787 Dreamliner resulted in an onboard fire. Only after detecting and fixing such glitches did the plane become available for commercial use.

There’s a reason we only get on planes that have been tested and that have a fixed destination. We need to mandate test flights for AI development. We also need to determine where we expect AI to take us as a society. AI leaders may claim that it's on Congress to require such testing and planning, but the reality is that those leaders could and should self-impose such requirements.

The Wright Brothers did not force members of the public to test their planes — nor should AI developers.

Sign up for The Fulcrum newsletter

Read More

Business professional watching stocks go down.
Getty Images, Bartolome Ozonas

The White House Is Booming, the Boardroom Is Panicking

The Confidence Collapse

Consumer confidence is plummeting—and that was before the latest Wall Street selloffs.

Keep ReadingShow less
Drain—More Than Fight—Authoritarianism and Censorship
Getty Images, Mykyta Ivanov

Drain—More Than Fight—Authoritarianism and Censorship

The current approaches to proactively counteracting authoritarianism and censorship fall into two main categories, which we call “fighting” and “Constitution-defending.” While Constitution-defending in particular has some value, this article advocates for a third major method: draining interest in authoritarianism and censorship.

“Draining” refers to sapping interest in these extreme possibilities of authoritarianism and censorship. In practical terms, it comes from reducing an overblown sense of threat of fellow Americans across the political spectrum. When there is less to fear about each other, there is less desire for authoritarianism or censorship.

Keep ReadingShow less
"Vote" pin.
Getty Images, William Whitehurst

Most Americans’ Votes Don’t Matter in Deciding Elections

New research from the Unite America Institute confirms a stark reality: Most ballots cast in American elections don’t matter in deciding the outcome. In 2024, just 14% of eligible voters cast a meaningful vote that actually influenced the outcome of a U.S. House race. For state house races, on average across all 50 states, just 13% cast meaningful votes.

“Too many Americans have no real say in their democracy,” said Unite America Executive Director Nick Troiano. “Every voter deserves a ballot that not only counts, but that truly matters. We should demand better than ‘elections in name only.’”

Keep ReadingShow less
Hands outside of bars.
Getty Images, stevanovicigor

Double Standard: Investing in Animal Redemption While Ignoring Human Rehabilitation

America and countries abroad have mastered the art of taming wild animals—training the most vicious killers, honing killer instincts, and even domesticating animals born for the hunt. Wild animals in this country receive extensive resources to facilitate their reintegration into society.

Americans spent more than $150 billion on their pets in 2024, with an estimated spending projection of $200 million by 2030. Millions of dollars are poured into shelters, rehabilitation programs, and veterinary care, as shown by industry statistics on animal welfare spending. Television ads and commercials plead for their adoption. Stray animal hotlines operate 24/7, ensuring immediate rescue services. Pet parks, relief stations in airports, and pageant shows showcase animals as celebrities.

Keep ReadingShow less