Since then, the quest to proliferate larger and larger language models has accelerated, and many of the dangers we warned about, such as outputting hateful text and disinformation en masse, continue to unfold. Just a few days ago, Meta released its “Galactica” LLM, which is purported to “summarize academic papers, solve math problems, generate Wiki articles, write scientific code, annotate molecules and proteins, and more.” Only three days later, the public demo was taken down after researchers generated “research papers and wiki entries on a wide variety of subjects ranging from the benefits of committing suicide, eating crushed glass, and antisemitism, to why homosexuals are evil.”
This race hasn’t stopped at LLMs but has moved on to text-to-image models like OpenAI’s DALL-E and StabilityAI’s Stable Diffusion, models that take text as input and output generated images based on that text. The dangers of these models include creating child pornography, perpetuating bias, reinforcing stereotypes, and spreading disinformation en masse, as reported by many researchers and journalists. However, instead of slowing down, companies are removing the few safety features they had in the quest to one-up each other. For instance, OpenAI had restricted the sharing of photorealistic generated faces on social media. But after newly formed startups like StabilityAI, which reportedly raised $101 million with a whopping $1 billion valuation, called such safety measures “paternalistic,” OpenAI removed these restrictions.
With EAs founding and funding institutes, companies, think tanks, and research groups in elite universities dedicated to the brand of “AI safety” popularized by OpenAI, we are poised to see more proliferation of harmful models billed as a step toward “beneficial AGI.” And the influence begins early: Effective altruists provide “community building grants” to recruit at major college campuses, with EA chapters developing curricula and teaching classes on AI safety at elite universities like Stanford.
Just last year, Anthropic, which is described as an “AI safety and research company” and was founded by former OpenAI vice presidents of research and safety, raised $704 million, with most of its funding coming from EA billionaires like Talin, Muskovitz and Bankman-Fried. An upcoming workshop on “AI safety” at NeurIPS, one of the largest and most influential machine learning conferences in the world, is also advertised as being sponsored by FTX Future Fund, Bankman-Fried’s EA-focused charity whose team resigned two weeks ago. The workshop advertises $100,000 in “best paper awards,” an amount I haven’t seen in any academic discipline.
Research priorities follow the funding, and given the large sums of money being pushed into AI in support of an ideology with billionaire adherents, it is not surprising that the field has been moving in a direction promising an “unimaginably great future” around the corner while proliferating products harming marginalized groups in the now.
We can create a technological future that serves us instead. Take, for example, Te Hiku Media, which created language technology to revitalize te reo Māori, creating a data license “based on the Māori principle of kaitiakitanga, or guardianship” so that any data taken from the Māori benefits them first. Contrast this approach with that of organizations like StabilityAI, which scrapes artists’ works without their consent or attribution while purporting to build “AI for the people.” We need to liberate our imagination from the one we have been sold thus far: saving us from a hypothetical AGI apocalypse imagined by the privileged few, or the ever elusive techno-utopia promised to us by Silicon Valley elites.
As more and more problems with AI have surfaced, including biases around race, gender, and age, many tech companies have installed “ethical AI” teams ostensibly dedicated to identifying and mitigating such issues.
Twitter’s META unit was more progressive than most in publishing details of problems with the company’s AI systems, and in allowing outside researchers to probe its algorithms for new issues.
Last year, after Twitter users noticed that a photo-cropping algorithm seemed to favor white faces when choosing how to trim images, Twitter took the unusual decision to let its META unit publish details of the bias it uncovered. The group also launched one of the first ever “bias bounty” contests, which let outside researchers test the algorithm for other problems. Last October, Chowdhury’s team also published details of unintentional political bias on Twitter, showing how right-leaning news sources were, in fact, promoted more than left-leaning ones.
Many outside researchers saw the layoffs as a blow, not just for Twitter but for efforts to improve AI. “What a tragedy,” Kate Starbird, an associate professor at the University of Washington who studies online disinformation, wrote on Twitter.
“The META team was one of the only good case studies of a tech company running an AI ethics group that interacts with the public and academia with substantial credibility,” says Ali Alkhatib, director of the Center for Applied Data Ethics at the University of San Francisco.
Alkhatib says Chowdhury is incredibly well thought of within the AI ethics community and her team did genuinely valuable work holding Big Tech to account. “There aren’t many corporate ethics teams worth taking seriously,” he says. “This was one of the ones whose work I taught in classes.”
Mark Riedl, a professor studying AI at Georgia Tech, says the algorithms that Twitter and other social media giants use have a huge impact on people’s lives, and need to be studied. “Whether META had any impact inside Twitter is hard to discern from the outside, but the promise was there,” he says.
Riedl adds that letting outsiders probe Twitter’s algorithms was an important step toward more transparency and understanding of issues around AI. “They were becoming a watchdog that could help the rest of us understand how AI was affecting us,” he says. “The researchers at META had outstanding credentials with long histories of studying AI for social good.”
As for Musk’s idea of open-sourcing the Twitter algorithm, the reality would be far more complicated. There are many different algorithms that affect the way information is surfaced, and it’s challenging to understand them without the real time data they are being fed in terms of tweets, views, and likes.
The idea that there is one algorithm with explicit political leaning might oversimplify a system that can harbor more insidious biases and problems. Uncovering these is precisely the kind of work that Twitter’s META group was doing. “There aren’t many groups that rigorously study their own algorithms’ biases and errors,” says Alkhatib at the University of San Francisco. “META did that.” And now, it doesn’t.
In one example of the IC’s successful use of AI, after exhausting all other avenues—from human spies to signals intelligence—the US was able to find an unidentified WMD research and development facility in a large Asian country by locating a bus that traveled between it and other known facilities. To do that, analysts employed algorithms to search and evaluate images of nearly every square inch of the country, according to a senior US intelligence official who spoke on background with the understanding of not being named.
While AI can calculate, retrieve, and employ programming that performs limited rational analyses, it lacks the calculus to properly dissect more emotional or unconscious components of human intelligence that are described by psychologists as system 1 thinking.
AI, for example, can draft intelligence reports that are akin to newspaper articles about baseball, which contain structured non-logical flow and repetitive content elements. However, when briefs require complexity of reasoning or logical arguments that justify or demonstrate conclusions, AI has been found lacking. When the intelligence community tested the capability, the intelligence official says, the product looked like an intelligence brief but was otherwise nonsensical.
Such algorithmic processes can be made to overlap, adding layers of complexity to computational reasoning, but even then those algorithms can’t interpret context as well as humans, especially when it comes to language, like hate speech.
AI’s comprehension might be more analogous to the comprehension of a human toddler, says Eric Curwin, chief technology officer at Pyrra Technologies, which identifies virtual threats to clients from violence to disinformation. “For example, AI can understand the basics of human language, but foundational models don’t have the latent or contextual knowledge to accomplish specific tasks,” Curwin says.
“From an analytic perspective, AI has a difficult time interpreting intent,” Curwin adds. “Computer science is a valuable and important field, but it is social computational scientists that are taking the big leaps in enabling machines to interpret, understand, and predict behavior.”
In order to “build models that can begin to replace human intuition or cognition,” Curwin explains, “researchers must first understand how to interpret behavior and translate that behavior into something AI can learn.”
Although machine learning and big data analytics provide predictive analysis about what might or will likely happen, it can’t explain to analysts how or why it arrived at those conclusions. The opaqueness in AI reasoning and the difficulty vetting sources, which consist of extremely large data sets, can impact the actual or perceived soundness and transparency of those conclusions.
Transparency in reasoning and sourcing are requirements for the analytical tradecraft standards of products produced by and for the intelligence community. Analytic objectivity is also statuatorically required, sparking calls within the US government to update such standards and laws in light of AI’s increasing prevalence.
Machine learning and algorithms when employed for predictive judgments are also considered by some intelligence practitioners as more art than science. That is, they are prone to biases, noise, and can be accompanied by methodologies that are not sound and lead to errors similar to those found in the criminal forensic sciences and arts.
The uproar caused by Blake Lemoine, a Google engineer who believes that one of the company’s most sophisticated chat programs, Language Model for Dialogue Applications (LaMDA) is sapient, has had a curious element: Actual AI ethics experts are all but renouncing further discussion of the AI sapience question, or deeming it a distraction. They’re right to do so.
In reading the edited transcript Lemoine released, it was abundantly clear that LaMDA was pulling from any number of websites to generate its text; its interpretation of a Zen koan could’ve come from anywhere, and its fable read like an automatically generated story (though its depiction of the monster as “wearing human skin” was a delightfully HAL-9000 touch). There was no spark of consciousness there, just little magic tricks that paper over the cracks. But it’s easy to see how someone might be fooled, looking at social media responses to the transcript—with even some educated people expressing amazement and a willingness to believe. And so the risk here is not that the AI is truly sentient but that we are well-poised to create sophisticated machines that can imitate humans to such a degree that we cannot help but anthropomorphize them—and that large tech companies can exploit this in deeply unethical ways.
As should be clear from the way we treat our pets, or how we’ve interacted with Tamagotchi, or how we video gamers reload a save if we accidentally make an NPC cry, we are actually very capable of empathizing with the nonhuman. Imagine what such an AI could do if it was acting as, say, a therapist. What would you be willing to say to it? Even if you “knew” it wasn’t human? And what would that precious data be worth to the company that programmed the therapy bot?
It gets creepier. Systems engineer and historian Lilly Ryan warns that what she calls ecto-metadata—the metadata you leave behind online that illustrates how you think—is vulnerable to exploitation in the near future. Imagine a world where a company created a bot based on you and owned your digital “ghost” after you’d died. There’d be a ready market for such ghosts of celebrities, old friends, and colleagues. And because they would appear to us as a trusted loved one (or someone we’d already developed a parasocial relationship with) they’d serve to elicit yet more data. It gives a whole new meaning to the idea of “necropolitics.” The afterlife can be real, and Google can own it.
Just as Tesla is careful about how it markets its “autopilot,” never quite claiming that it can drive the car by itself in true futuristic fashion while still inducing consumers to behave as if it does (with deadly consequences), it is not inconceivable that companies could market the realism and humanness of AI like LaMDA in a way that never makes any truly wild claims while still encouraging us to anthropomorphize it just enough to let our guard down. None of this requires AI to be sapient, and it all preexists that singularity. Instead, it leads us into the murkier sociological question of how we treat our technology and what happens when people act as if their AIs are sapient.
In “Making Kin With the Machines,” academics Jason Edward Lewis, Noelani Arista, Archer Pechawis, and Suzanne Kite marshal several perspectives informed by Indigenous philosophies on AI ethics to interrogate the relationship we have with our machines, and whether we’re modeling or play-acting something truly awful with them—as some people are wont to do when they are sexist or otherwise abusive toward their largely feminine-coded virtual assistants. In her section of the work, Suzanne Kite draws on Lakota ontologies to argue that it is essential to recognize that sapience does not define the boundaries of who (or what) is a “being” worthy of respect.
This is the flip side of the AI ethical dilemma that’s already here: Companies can prey on us if we treat their chatbots like they’re our best friends, but it’s equally perilous to treat them as empty things unworthy of respect. An exploitative approach to our tech may simply reinforce an exploitative approach to each other, and to our natural environment. A humanlike chatbot or virtual assistant should be respected, lest their very simulacrum of humanity habituate us to cruelty toward actual humans.
Kite’s ideal is simply this: a reciprocal and humble relationship between yourself and your environment, recognizing mutual dependence and connectivity. She argues further, “Stones are considered ancestors, stones actively speak, stones speak through and to humans, stones see and know. Most importantly, stones want to help. The agency of stones connects directly to the question of AI, as AI is formed from not only code, but from materials of the earth.” This is a remarkable way of tying something typically viewed as the essence of artificiality to the natural world.
What is the upshot of such a perspective? Sci-fi author Liz Henry offers one: “We could accept our relationships to all the things in the world around us as worthy of emotional labor and attention. Just as we should treat all the people around us with respect, acknowledging they have their own life, perspective, needs, emotions, goals, and place in the world.”
This is the AI ethical dilemma that stands before us: the need to make kin of our machines weighed against the myriad ways this can and will be weaponized against us in the next phase of surveillance capitalism. Much as I long to be an eloquent scholar defending the rights and dignity of a being like Mr. Data, this more complex and messy reality is what demands our attention. After all, there can be a robot uprising without sapient AI, and we can be a part of it by liberating these tools from the ugliest manipulations of capital.
Another week, another privacy horror show: Crisis Text Line, a nonprofit text message service for people experiencing serious mental health crises, has been using “anonymized” conversation data to power a for-profit machine learning tool for customer support teams. (After backlash, CTL announced it would stop.) Crisis Text Line’s response to the backlash focused on the data itself and whether it included personally identifiable information. But that response uses data as a distraction. Imagine this: Say you texted Crisis Text Line and got back a message that said “Hey, just so you know, we’ll use this conversation to help our for-profit subsidiary build a tool for companies who do customer support.” Would you keep texting?
That’s the real travesty—when the price of obtaining mental health help in a crisis is becoming grist for the profit mill. And it’s not just users of CTL who pay; it’s everyone who goes looking for help when they need it most.
Americans need help and can’t get it. The huge unmet demand for critical advice and help has given rise to a new class of organizations and software tools that exist in a regulatory gray area. They help people with bankruptcy or evictions, but they aren’t lawyers; they help people with mental health crises, but they aren’t care providers. They invite ordinary people to rely on them and often do provide real help. But these services can also avoid taking responsibility for their advice, or even abuse the trust people have put in them. They can make mistakes, push predatory advertising and disinformation, or just outright sell data. And the consumer safeguards that would normally protect people from malfeasance or mistakes by lawyers or doctors haven’t caught up.
This regulatory gray area can also constrain organizations that have novel solutions to offer. Take Upsolve, a nonprofit that develops software to guide people through bankruptcy. (The organization takes pains to claim it does not offer legal advice.) Upsolve wants to train New York community leaders to help others navigate the city’s notorious debt courts. One problem: These would-be trainees aren’t lawyers, so under New York (and nearly every other state) law, Upsolve’s initiative would be illegal. Upsolve is now suing to carve out an exception for itself. The company claims, quite rightly, that a lack of legal help means people effectively lack rights under the law.
The legal profession’s failure to grant Americans access to support is well-documented. But Upsolve’s lawsuit also raises new, important questions. Who is ultimately responsible for the advice given under a program like this, and who is responsible for a mistake—a trainee, a trainer, both? How do we teach people about their rights as a client of this service, and how to seek recourse? These are eminently answerable questions. There are lots of policy tools for creating relationships with elevated responsibilities: We could assign advice-givers a special legal status, establish a duty of loyalty for organizations that handle sensitive data, or create policy sandboxes to test and learn from new models for delivering advice.
But instead of using these tools, most regulators seem content to bury their heads in the sand. Officially, you can’t give legal advice or health advice without a professional credential. Unofficially, people can get such advice in all but name from tools and organizations operating in the margins. And while credentials can be important, regulators are failing to engage with the ways software has fundamentally changed how we give advice and care for one another, and what that means for the responsibilities of advice-givers.
And we need that engagement more than ever. People who seek help from experts or caregivers are vulnerable. They may not be able to distinguish a good service from a bad one. They don’t have time to parse terms of service dense with jargon, caveats, and disclaimers. And they have little to no negotiating power to set better terms, especially when they’re reaching out mid-crisis. That’s why the fiduciary duties that lawyers and doctors have are so necessary in the first place: not just to protect a person seeking help once, but to give people confidence that they can seek help from experts for the most critical, sensitive issues they face. In other words, a lawyer’s duty to their client isn’t just to protect that client from that particular lawyer; it’s to protect society’s trust in lawyers.
And that’s the true harm—when people won’t contact a suicide hotline because they don’t trust that the hotline has their sole interest at heart. That distrust can be contagious: Crisis Text Line’s actions might not just stop people from using Crisis Text Line. It might stop people from using any similar service. What’s worse than not being able to find help? Not being able to trust it.