Republicans just got a painful reminder that political stunts can backfire.

On Wednesday, a judge in the deep-red state of Wyoming temporarily blocked a state law that would make performing nearly any abortion in that state a felony. She relied on a 2012 amendment to the state constitution that was intended to spite then-President Barack Obama.

Obama’s early years in office were marred by a scorched-earth political campaign Republicans wielded to try to thwart what became the Affordable Care Act. Obamacare’s opponents warned of a “government takeover of health care” that would strip many Americans of their ability to make their own health decisions.

Many of these allegations were downright ludicrous, such as former Alaska Gov. Sarah Palin’s (R) false claim that Obama’s health bill would require “my baby with Down Syndrome ... to stand in front of Obama’s ‘death panel’ so his bureaucrats can decide, based on a subjective judgment of their ‘level of productivity in society’ whether they are worthy of health care.”

These attacks did not succeed. The bill became law, and Obamacare is popular now that it has been in full effect for nearly a decade without anyone being forced to stand before a death panel. But there is at least one lasting legacy of these attempts to characterize the Affordable Care Act as an attack on patients’ right to decide whether and when to seek health treatments.

In many states, opponents of Obamacare effectively took the GOP’s talking points and turned them into state constitutional amendments protecting patients’ ability to obtain health care that the government might not want them to have. Wyoming’s amendment, for example, provides that “each competent adult shall have the right to make his or her own health care decisions.”

According to Quinn Yeargain, a law professor at Widener University, similar amendments are on the books in several other states.

It remains to be seen whether the highest courts in these states, some of which are extremely conservative, will ultimately agree that these anti-Obamacare amendments prohibit abortion bans. And, in at least some cases, the amendments contain language that could mitigate their impact. Wyoming’s amendment, for example, also provides that, under certain circumstances, the state legislature may “determine reasonable and necessary restrictions on the rights granted” by the health care amendment.

But abortion advocates have had two early successes: the Wyoming judge’s order temporarily blocking that state’s abortion ban, and a similar decision by a trial judge in Ohio.

The Wyoming abortion rights litigation, briefly explained

Wyoming district court Judge Melissa Owens’s Wednesday decision temporarily halting her state’s abortion ban is the second time she intervened to prevent this ban from going into effect. Wyoming’s abortion ban is quite strict, although it does provide exceptions for rape, incest, or when either a pregnant patient or the fetus has certain medical conditions.

Last summer, shortly after the Supreme Court’s decision overruling Roe v. Wade, an array of patients, doctors, and nonprofit groups brought a suit arguing that Wyoming’s abortion ban violated the state’s constitutional provision protecting each adult’s right to individual health care decisions. That case is known as Johnson v. Wyoming.

Judge Owens handed down a decision in August halting the law. Among other things, she rejected the state’s argument that the health care amendment was “only adopted to push back against the Affordable Care Act,” and should not be construed to protect abortion rights.

Regardless of the political circumstances that led to this amendment being written into the state constitution, Owens reasoned that the amendment “unambiguously provides competent Wyoming citizens with the right to make their own health care decisions,” and she was bound by that unambiguous text. “A court,” she wrote, “is not at liberty to assume that the Wyoming voters who adopted” the amendment “did not understand the force of language in the provision.”

Just as significantly, Owens construed the amendment to give people in Wyoming a “fundamental right” to make their own health care decisions, including the decision to seek an abortion. This designation matters because fundamental rights can only be abridged when the state seeks to advance a “compelling state interest” and when it uses the “least intrusive” means to do so.

Thus, even though the amendment permits the state legislature to impose “reasonable and necessary restrictions” on individual’s health choices, Owens concluded that Wyoming’s broad ban on abortion access sweeps too far because it intrudes into pregnant patients’ health care decisions even when a “fetus has a genetic abnormality that is incompatible with life.” (The state has since amended its law to permit abortions when “there is a substantial likelihood that the unborn baby has a lethal fetal anomaly,” a change that could undermine Owens’s legal reasoning.)

There is precedent for Owens’s conclusion that this Wyoming health care amendment establishes a fundamental right that the legislature may only abridge under very limited circumstances, even though that same amendment gives the legislature some authority to enact laws. The US Constitution’s 14th Amendment has long been construed to protect many fundamental rights, such as the right to marry or the right to choose your own sexual partners. But the 14th Amendment also contains language permitting Congress to enforce its provisions “by appropriate legislation.”

Nevertheless, the fact that the 14th Amendment permits Congress to enact laws it deems “appropriate” typically does not permit Congress to abridge the fundamental rights it guarantees.

In response to Owens’s August decision blocking the state’s abortion ban, the state legislature enacted a new law decreeing that abortion “is not health care” and thus is not protected by the state constitution. Owens’s Wednesday order blocked that law as well, declaring that “the legislature cannot make an end run around” around a constitutional amendment, and that it is up to the courts to decide whether abortion meets the state constitution’s definition of “health care.”

Yet, while the state legislature appears eager to restore the state’s abortion ban, the Wyoming Supreme Court has thus far resisted the urge to rush in and overrule Owens. Last December, after a case reached the state Supreme Court that it could have used to reject Judge Owens’s reading of the state constitution, Wyoming’s justices chose instead not to decide that case. That left Owens’s August order in effect.

So, while there are plausible legal arguments on either side of this dispute, there appears to be a real chance that the state’s highest court will agree with Owens if and when they weigh in on whether the state constitution protects abortion. If the state Supreme Court shared the legislature’s view that abortion must be banned in Wyoming, it could have intervened last winter.

Could anti-Obamacare amendments protect abortion rights in other states?

At least one other state court, in Ohio, relied on that state’s anti-Obamacare amendment in an opinion temporarily blocking a law that bans nearly all abortions after the sixth week of pregnancy. That 2022 decision, in a case known as Preterm-Cleveland v. Yost, argued that a few provisions of the state constitution, including the state’s health care amendment, work together to protect abortion rights.

Last December, a state appeals court decided that the trial court’s order in Preterm-Cleveland may remain in effect, at least for now.

Ohio’s amendment provides that no state law “shall prohibit the purchase or sale of health care or health insurance.” Nor may it “impose a penalty or fine for the sale or purchase of health care or health insurance.” Thus, as long as a patient seeking an abortion pays for that treatment, the Ohio amendment appears to provide very robust protection to abortion rights.

Like the Wyoming amendment, Ohio’s permits the legislature to enact some restrictions on the right to purchase health care but the Ohio amendment uses less expansive language to describe when such restrictions are allowed — though one provision of the Ohio amendment does permit state laws that are “calculated to deter fraud or punish wrongdoing in the health care industry.” An abortion opponent would no doubt argue that abortions are themselves a form of “wrongdoing.”

In any event, the Ohio Supreme Court has a 4-3 Republican majority. So there’s no guarantee that the state’s justices will agree with the trial court’s ruling and allow abortion to remain legal in Ohio.

(Until recently, the swing vote on the Ohio Supreme Court was held by Chief Justice Maureen O’Connor, a relatively moderate Republican. But O’Connor recently retired and the Court’s new majority hasn’t developed much of a record. So it is difficult for a lawyer to assess with certainty how it is likely to rule on a case like Preterm-Cleveland.)

But what about other states that enacted health care amendments as a statement of defiance against Obamacare? The short answer is that a lawsuit seeking to protect abortion rights in these states would turn on the same questions that are in play in Wyoming and Ohio: What does the state’s health care amendment actually say? And who controls the state Supreme Court?

Alabama’s amendment, for example, is unlikely to help abortion advocates very much, even setting aside the fact that Alabama’s Supreme Court is dominated by Republicans. That’s because Alabama’s amendment primarily prohibits the state from requiring “any person, employer, or health care provider to participate in any health care system.” That language cannot reasonably be construed to protect abortion rights.

Other states, including Arizona, Missouri, and Oklahoma, enacted similar amendments preventing the state government from compelling individuals to “participate in any health care system.” These amendments are also unlikely to help proponents of abortion rights.

So this largely forgotten legacy of a failed Republican effort to spite Obamacare is only likely to matter in a very small number of states. And it may not even have a lasting impact in Wyoming and Ohio, depending on how their state Supreme Courts rule on whether the state constitution protects abortion.

For the moment, however, the Obama-era amendments writing anti-Obamacare talking points into two state constitutions have proved to be a thorn in the side of Republicans who hope to ban abortions. Let that be a lesson that a state constitution is a foolish thing to change for the sake of a political stunt.

If you don’t believe the rushed launch of AI chatbots by Big Tech has an extremely strong chance of degrading the web’s information ecosystem, consider the following:

Right now,* if you ask Microsoft’s Bing chatbot if Google’s Bard chatbot has been shut down, it says yes, citing as evidence a news article that discusses a tweet in which a user asked Bard when it would be shut down and Bard said it already had, itself citing a comment from Hacker News in which someone joked about this happening, and someone else used ChatGPT to write fake news coverage about the event.

(*I say “right now” because in the time between starting and finishing writing this story, Bing changed its answer and now correctly replies that Bard is still live. You can interpret this as showing that these systems are, at least, fixable or that they are so infinitely malleable that it’s impossible to even consistently report their mistakes.)

Image: The Verge

But if reading all that made your head hurt, it should — and in more ways than one.

What we have here is an early sign we’re stumbling into a massive game of AI misinformation telephone, in which chatbots are unable to gauge reliable news sources, misread stories about themselves, and misreport on their own capabilities. In this case, the whole thing started because of a single joke comment on Hacker News. Imagine what you could do if you wanted these systems to fail.

It’s a laughable situation but one with potentially serious consequences. Given the inability of AI language models to reliably sort fact from fiction, their launch online threatens to unleash a rotten trail of misinformation and mistrust across the web, a miasma that is impossible to map completely or debunk authoritatively. All because Microsoft, Google, and OpenAI have decided that market share is more important than safety.

These companies can put as many disclaimers as they like on their chatbots — telling us they’re “experiments,” “collaborations,” and definitely not search engines — but it’s a flimsy defense. We know how people use these systems, and we’ve already seen how they spread misinformation, whether inventing new stories that were never written or telling people about books that don’t exist. And now, they’re citing one another’s mistakes, too.

Click here to go see the bonus panel!

Hovertext:
Also you'll want to call this tech support number your father called before you.

This is probably the right place to mention we made the thing people kept asking for.

Nearly 90% of the multibillion-dollar federal lobbying apparatus in the United States serves corporate interests. In some cases, the objective of that money is obvious. Google pours millions into lobbying on bills related to antitrust regulation. Big energy companies expect action whenever there is a move to end drilling leases for federal lands, in exchange for the tens of millions they contribute to congressional reelection campaigns.

But lobbying strategies are not always so blunt, and the interests involved are not always so obvious. Consider, for example, a 2013 Massachusetts bill that tried to restrict the commercial use of data collected from K-12 students using services accessed via the internet. The bill appealed to many privacy-conscious education advocates, and appropriately so. But behind the justification of protecting students lay a market-altering policy: the bill was introduced at the behest of Microsoft lobbyists, in an effort to exclude Google Docs from classrooms.

What would happen if such legal-but-sneaky strategies for tilting the rules in favor of one group over another become more widespread and effective? We can see hints of an answer in the remarkable pace at which artificial-intelligence tools for everything from writing to graphic design are being developed and improved. And the unavoidable conclusion is that AI will make lobbying more guileful, and perhaps more successful.

It turns out there is a natural opening for this technology: microlegislation.

“Microlegislation” is a term for small pieces of proposed law that cater—sometimes unexpectedly—to narrow interests. Political scientist Amy McKay coined the term. She studied the 564 amendments to the Affordable Care Act (“Obamacare”) considered by the Senate Finance Committee in 2009, as well as the positions of 866 lobbying groups and their campaign contributions. She documented instances where lobbyist comments—on health-care research, vaccine services, and other provisions—were translated directly into microlegislation in the form of amendments. And she found that those groups’ financial contributions to specific senators on the committee increased the amendments’ chances of passing.

Her finding that lobbying works was no surprise. More important, McKay’s work demonstrated that computer models can predict the likely fate of proposed legislative amendments, as well as the paths by which lobbyists can most effectively secure their desired outcomes. And that turns out to be a critical piece of creating an AI lobbyist.

Lobbying has long been part of the give-and-take among human policymakers and advocates working to balance their competing interests. The danger of microlegislation—a danger greatly exacerbated by AI—is that it can be used in a way that makes it difficult to figure out who the legislation truly benefits.

Another word for a strategy like this is a “hack.” Hacks follow the rules of a system but subvert their intent. Hacking is often associated with computer systems, but the concept is also applicable to social systems like financial markets, tax codes, and legislative processes.

While the idea of monied interests incorporating AI assistive technologies into their lobbying remains hypothetical, specific machine-learning technologies exist today that would enable them to do so. We should expect these techniques to get better and their utilization to grow, just as we’ve seen in so many other domains.

Here’s how it might work.

Crafting an AI microlegislator

To make microlegislation, machine-learning systems must be able to uncover the smallest modification that could be made to a bill or existing law that would make the biggest impact on a narrow interest.

There are three basic challenges involved. First, you must create a policy proposal—small suggested changes to legal text—and anticipate whether or not a human reader would recognize the alteration as substantive. This is important; a change that isn’t detectable is more likely to pass without controversy. Second, you need to do an impact assessment to project the implications of that change for the short- or long-range financial interests of companies. Third, you need a lobbying strategizer to identify what levers of power to pull to get the best proposal into law.

Existing AI tools can tackle all three of these.

The first step, the policy proposal, leverages the core function of generative AI. Large language models, the sort that have been used for general-purpose chatbots such as ChatGPT, can easily be adapted to write like a native in different specialized domains after seeing a relatively small number of examples. This process is called fine-tuning. For example, a model “pre-trained” on a large library of generic text samples from books and the internet can be “fine-tuned” to work effectively on medical literature, computer science papers, and product reviews.

Given this flexibility and capacity for adaptation, a large language model could be fine-tuned to produce draft legislative texts, given a data set of previously offered amendments and the bills they were associated with. Training data is available. At the federal level, it’s provided by the US Government Publishing Office, and there are already tools for downloading and interacting with it. Most other jurisdictions provide similar data feeds, and there are even convenient assemblages of that data.

Meanwhile, large language models like the one underlying ChatGPT are routinely used for summarizing long, complex documents (even laws and computer code) to capture the essential points, and they are optimized to match human expectations. This capability could allow an AI assistant to automatically predict how detectable the true effect of a policy insertion may be to a human reader.

Today, it can take a highly paid team of human lobbyists days or weeks to generate and analyze alternative pieces of microlegislation on behalf of a client. With AI assistance, that could be done instantaneously and cheaply. This opens the door to dramatic increases in the scope of this kind of microlegislating, with a potential to scale across any number of bills in any jurisdiction.

Teaching machines to assess impact

Impact assessment is more complicated. There is a rich series of methods for quantifying the predicted outcome of a decision or policy, and then also optimizing the return under that model. This kind of approach goes by different names in different circles—mathematical programming in management science, utility maximization in economics, and rational design in the life sciences.

To train an AI to do this, we would need to specify some way to calculate the benefit to different parties as a result of a policy choice. That could mean estimating the financial return to different companies under a few different scenarios of taxation or regulation. Economists are skilled at building risk models like this, and companies are already required to formulate and disclose regulatory compliance risk factors to investors. Such a mathematical model could translate directly into a reward function, a grading system that could provide feedback for the model used to create policy proposals and direct the process of training it.

The real challenge in impact assessment for generative AI models would be to parse the textual output of a model like ChatGPT in terms that an economic model could readily use. Automating this would require extracting structured financial information from the draft amendment or any legalese surrounding it. This kind of information extraction, too, is an area where AI has a long history; for example, AI systems have been trained to recognize clinical details in doctors’ notes. Early indications are that large language models are fairly good at recognizing financial information in texts such as investor call transcripts. While it remains an open challenge in the field, they may even be capable of writing out multi-step plans based on descriptions in free text.

Machines as strategists

The last piece of the puzzle is a lobbying strategizer to figure out what actions to take to convince lawmakers to adopt the amendment.

Passing legislation requires a keen understanding of the complex interrelated networks of legislative offices, outside groups, executive agencies, and other stakeholders vying to serve their own interests. Each actor in this network has a baseline perspective and different factors that influence that point of view. For example, a legislator may be moved by seeing an allied stakeholder take a firm position, or by a negative news story, or by a campaign contribution.

It turns out that AI developers are very experienced at modeling these kinds of networks. Machine-learning models for network graphs have been built, refined, improved, and iterated by hundreds of researchers working on incredibly diverse problems: lidar scans used to guide self-driving cars, the chemical functions of molecular structures, the capture of motion in actors’ joints for computer graphics, behaviors in social networks, and more.

In the context of AI-assisted lobbying, political actors like legislators and lobbyists are nodes on a graph, just like users in a social network. Relations between them are graph edges, like social connections. Information can be passed along those edges, like messages sent to a friend or campaign contributions made to a member. AI models can use past examples to learn to estimate how that information changes the network. Calculating the likelihood that a campaign contribution of a given size will flip a legislator’s vote on an amendment is one application.

McKay’s work has already shown us that there are significant, predictable relationships between these actions and the outcomes of legislation, and that the work of discovering those can be automated. Others have shown that graphs of neural network models like those described above can be applied to political systems. The full-scale use of these technologies to guide lobbying strategy is theoretical, but plausible.

Put together, these three components could create an automatic system for generating profitable microlegislation. The policy proposal system would create millions, even billions, of possible amendments. The impact assessor would identify the few that promise to be most profitable to the client. And the lobbying strategy tool would produce a blueprint for getting them passed.

What remains is for human lobbyists to walk the floors of the Capitol or state house, and perhaps supply some cash to grease the wheels. These final two aspects of lobbying—access and financing—cannot be supplied by the AI tools we envision. This suggests that lobbying will continue to primarily benefit those who are already influential and wealthy, and AI assistance will amplify their existing advantages.

The transformative benefit that AI offers to lobbyists and their clients is scale. While individual lobbyists tend to focus on the federal level or a single state, with AI assistance they could more easily infiltrate a large number of state-level (or even local-level) law-making bodies and elections. At that level, where the average cost of a seat is measured in the tens of thousands of dollars instead of millions, a single donor can wield a lot of influence—if automation makes it possible to coordinate lobbying across districts.

How to stop them

When it comes to combating the potentially adverse effects of assistive AI, the first response always seems to be to try to detect whether or not content was AI-generated. We could imagine a defensive AI that detects anomalous lobbyist spending associated with amendments that benefit the contributing group. But by then, the damage might already be done.

In general, methods for detecting the work of AI tend not to keep pace with its ability to generate convincing content. And these strategies won’t be implemented by AIs alone. The lobbyists will still be humans who take the results of an AI microlegislator and further refine the computer’s strategies. These hybrid human-AI systems will not be detectable from their output.

But the good news is: the same strategies that have long been used to combat misbehavior by human lobbyists can still be effective when those lobbyists get an AI assist. We don’t need to reinvent our democracy to stave off the worst risks of AI; we just need to more fully implement long-standing ideals.

First, we should reduce the dependence of legislatures on monolithic, multi-thousand-page omnibus bills voted on under deadline. This style of legislating exploded in the 1980s and 1990s and continues through to the most recent federal budget bill. Notwithstanding their legitimate benefits to the political system, omnibus bills present an obvious and proven vehicle for inserting unnoticed provisions that may later surprise the same legislators who approved them.

The issue is not that individual legislators need more time to read and understand each bill (that isn’t realistic or even necessary). It’s that omnibus bills must pass. There is an imperative to pass a federal budget bill, and so the capacity to push back on individual provisions that may seem deleterious (or just impertinent) to any particular group is small. Bills that are too big to fail are ripe for hacking by microlegislation.

Moreover, the incentive for legislators to introduce microlegislation catering to a narrow interest is greater if the threat of exposure is lower. To strengthen the threat of exposure for misbehaving legislative sponsors, bills should focus more tightly on individual substantive areas and, after the introduction of amendments, allow more time before the committee and floor votes. During this time, we should encourage public review and testimony to provide greater oversight.

Second, we should strengthen disclosure requirements on lobbyists, whether they’re entirely human or AI-assisted. State laws regarding lobbying disclosure are a hodgepodge. North Dakota, for example, only requires lobbying reports to be filed annually, so that by the time a disclosure is made, the policy is likely already decided. A lobbying disclosure scorecard created by Open Secrets, a group researching the influence of money in US politics, tracks nine states that do not even require lobbyists to report their compensation.

Ideally, it would be great for the public to see all communication between lobbyists and legislators, whether it takes the form of a proposed amendment or not. Absent that, let’s give the public the benefit of reviewing what lobbyists are lobbying for—and why. Lobbying is traditionally an activity that happens behind closed doors. Right now, many states reinforce that: they actually exempt testimony delivered publicly to a legislature from being reported as lobbying.

In those jurisdictions, if you reveal your position to the public, you’re no longer lobbying. Let’s do the inverse: require lobbyists to reveal their positions on issues. Some jurisdictions already require a statement of position (a ‘yea’ or ‘nay’) from registered lobbyists. And in most (but not all) states, you could make a public records request regarding meetings held with a state legislator and hope to get something substantive back. But we can expect more—lobbyists could be required to proactively publish, within a few days, a brief summary of what they demanded of policymakers during meetings and why they believe it’s in the general interest.

We can’t rely on corporations to be forthcoming and wholly honest about the reasons behind their lobbying positions. But having them on the record about their intentions would at least provide a baseline for accountability.

Finally, consider the role AI assistive technologies may have on lobbying firms themselves and the labor market for lobbyists. Many observers are rightfully concerned about the possibility of AI replacing or devaluing the human labor it automates. If the automating potential of AI ends up commodifying the work of political strategizing and message development, it may indeed put some professionals on K Street out of work.

But don’t expect that to disrupt the careers of the most astronomically compensated lobbyists: former members Congress and other insiders who have passed through the revolving door. There is no shortage of reform ideas for limiting the ability of government officials turned lobbyists to sell access to their colleagues still in government, and they should be adopted and—equally important—maintained and enforced in successive Congresses and administrations.

None of these solutions are really original, specific to the threats posed by AI, or even predominantly focused on microlegislation—and that’s the point. Good governance should and can be robust to threats from a variety of techniques and actors.

But what makes the risks posed by AI especially pressing now is how fast the field is developing. We expect the scale, strategies, and effectiveness of humans engaged in lobbying to evolve over years and decades. Advancements in AI, meanwhile, seem to be making impressive breakthroughs at a much faster pace—and it’s still accelerating.

The legislative process is a constant struggle between parties trying to control the rules of our society as they are updated, rewritten, and expanded at the federal, state, and local levels. Lobbying is an important tool for balancing various interests through our system. If it’s well-regulated, perhaps lobbying can support policymakers in making equitable decisions on behalf of us all.

This article was co-written with Nathan E. Sanders and originally appeared in MIT Technology Review.

Issue

Google’s Discover page, which is the leftmost page on many Android phones, can be a useful source of news, or it can be an irritating list of clickbait headlines and repetitive articles. And because it is an intrinsic part of the OS on many phones, unless you actually installed it yourself, you probably can’t uninstall it.

You can try to adjust the content by selecting the two dots at the bottom right of each article and letting the algorithm know you’re not interested, but that doesn’t always work. Or you can simply get rid of it.

Quick fix

Long-press on any homepage screen until you get a pop-up menu. Select Home settings and toggle off Swipe to access Google app. And that’s it — the Discover page is gone.

The full story

The Discover page uses an algorithm based on your usage to exhibit a listing of various articles pulled from the web that it thinks will interest you. However, even if you let it know what you’re interested or not interested in by using that two-dot More icon at the bottom right of each article, the list can still become packed with pushy headlines, repetitive blogs, or just plain wrong stuff.

If that is what’s happened on your phone and you just want to get rid of the whole thing, then follow the directions above. After that, you will no longer be able to swipe right from your initial homepage to the Discover page — it will simply be gone. (You can, of course, get it back by following the same directions and toggling Swipe to access Google app back on.)

Create a blank Discover page

There is another option if you don’t want to completely get rid of the Discover page but are tired of seeing that long list of headlines.

• From your Discover page or your Google app, tap your personalized icon in the top-right corner.
• Select Settings > General.
• Toggle Discover off.

You will now find your leftmost homepage is blank instead of filled with a long list of unwanted content. Well, almost blank — there will be a Turn on Discover button in case you change your mind.

Remove Discover from your Chrome app

Oh, and you may also find that the Discover page has invaded your phone’s Chrome app so that you see the Discover feed just below your recent search results every time you start a new tab. Luckily, it’s easy to get rid of.

• Start a new tab on your phone’s Chrome app, and tap the gear icon just below your search results.
• Tap the Turn off option on the drop-down menu.

If you don’t mind having Discover on your Chrome tabs but just don’t like the content you’re getting, tap the Manage option, and you’ll be offered several ways to try to tame the feed.

Enlarge (credit: dasilvafa)

Ever find yourself staring at a tricky coding problem and thinking, “shit”?

If those thoughts make their way into your code or the associated comments, you’re in good company. When undergraduate student Jan Strehmel from Karlsruhe Institute of Technology analyzed open source code written in the programming language C, he found no shortage of obscenity. While that might be expected, Strehmel’s overall finding might not be: The average quality of code containing swears was significantly higher than the average quality of code that did not.

“The results are quite surprising!” Strehmel said. Programmers and scientists may have a lot of follow-up questions. Are the researchers sure there aren’t certain profanity-prone programmers skewing the results? What about other programming languages? And, most importantly, why would swears correlate with high-quality code? The work is ongoing, but even without all the answers, one thing’s for sure: Strehmel just wrote one hell of a bachelor’s thesis.

Bad words, good code

Strehmel’s supervisor, Bioinformatician Alexandros Stamatakis, started wondering how swears affect code quality after a lab member showed him a graph of the prevalence of swears in various versions of the code underlying Linux. Stamatakis realized he had the perfect tool for asking whether profanity correlates with the quality of code. A program called SoftWipe, developed by his lab, measures adherence to coding standards, such as the use of quality checks and a simple code structure.

To investigate, Strehmel pulled around 3,800 examples of code containing swears, along with 7,600 examples of code that did not, from GitHub. SoftWipe revealed that on average, code containing swears scored about half a point higher on its 10- point scale of code quality than code that did not. “My reaction was that this is cool!” Stamatakis said. He frequently finds himself swearing at his own code, although he tends not to document his outbursts in text. Nonetheless, he wonders if his past curses may help his career progress: “Maybe that has helped me to become a full professor!” he said.

Psychologists have long known that swearing can relieve pain, increase physical performance, and help people shape their personas. In fact, cognitive psychologist Benjamin Bergen from the University of California San Diego—author of the book, What the F: What Swearing Reveals About Our Language, Our Brains, and Ourselves—makes a point to swear once during every college lecture he teaches (in a way that’s unlikely to offend the class) because there’s evidence that profanity, when used strategically, may increase student engagement.

But the link between swearing and code quality has not been examined before, as far as Bergen knows, and the suggestion that there’s a connection is a “very exciting, interesting idea,” he said.

The power of personality

Programmers who swear may be more emotionally engaged with their work than those who don’t, Bergen hypothesized, which could lead them to produce higher-quality products. Alternatively, programmers may include profanity to amuse or shock people who read their code—and if they expect their code to be read, they may put extra effort into it. It’s likely that swearing is a “symptom of something deeper going on,” Bergen said, and he’d like to see future work focus on the underlying cause of the association.

Software engineer Greg Wilson, who now works at the biotech company Deep Genomics, isn’t surprised to see coders’ personalities entering their work through their word choices. Wilson co-founded an organization called The Carpentries that teaches scientists to become good coders and says, “I don’t know anybody who’s good at anything who leaves themselves out of it.”

Wilson is excited to see researchers tackling the question of what makes code good, although Strehmel’s results are preliminary. Coders lag behind other disciplines in terms of how they evaluate their own work, he says. Unlike architects, who have nuanced ways of describing why a building is beautiful, programmers “can say that something is an elegant solution, and then we run out of words.”

He does worry about the impacts that profanity can have if it appears directed at junior programmers, however. Aggressive language has been cited as one factor that discourages people—especially those from groups that are marginalized in STEM—from continuing to work in software engineering. Strehmel and Stamatakis came across the occasional slurs in the code they analyzed, and they agree that there are lines programmers shouldn’t cross. At a certain point, “it stops being funny,” Stamatakis said.

Overall, however, the researchers are enjoying their work, and they have a long list of experiments planned to shore up the results and glean additional insight. When they’re ready to release their final product, Wilson is looking forward to seeing the commit message. He imagines it reading, “holy shit, it worked!”

Saima Sidik is a freelance science writer based in Somerville, Massachusetts. When she’s not writing, she enjoys biking around the city, learning photography, and practicing taekwondo.

Read Comments