Be the algorithm
Now more than ever, the “truth” is what we make it.
In a recent essay about why he retired from writing for the New York Times, economist Paul Krugman wrote the following:
“It became clear to me that the management I was dealing with didn’t understand the difference between having an opinion and having an informed, factually sourced opinion.”
Krugman shared this observation as something of a sidebar to his broader explanation for leaving the Times. But I want to elevate it because I think this problem, of our media pushing (and being rewarded for) sexy opinions over factual and logically sound ones, is a big part of how we have arrived in the dysfunctional and highly manipulated information ecosystem we have today.
My mentor, Craig Gerlach, is a man of many great turns of phrase. “Slipperier than deer guts on a door knob” is my all time favorite. But the one I now find myself using regularly is “seductively cogent but empirically vacant.” He would pull out this gem whenever confronted with some line of thinking or theory that felt sexy, but fell apart if you looked at it too long. Sexy because it was contrarian, perhaps. Sexy because it “slayed a sacred cow”. Sexy because it proved an expert wrong.
In other words, it was his way of saying that an argument was, “all hat no cattle” (another in Craig’s regular rotation).
I can tell you it feels good as a young and aspiring thinker/writer/scientist when you think you think you have an argument cocked and ready that will out-argue the experts. And it pays, both figuratively and literally. The world of academia rewards those who offer new ideas that upturn old ones. Novelty is the currency of many academic journals, and in crowded research fields, standing out is how you get ahead.
Academia, for what its worth, seems to be catching on to how perverse incentives regarding novelty are undermining the enterprise of science. But popular media, both traditional and social, is fully overwhelmed by the economic incentives of hot takes.
“If it bleeds, it leads” has long been part of the journalistic zeitgest, and journalistic ethics around fact checking and sourcing are relatively new in the profession’s history. But nevertheless, for a short period of time we could generally trust that editors and writers and reporters would check and double check their work. And we could trust that when they were reporting on a complex issue, they sought out experts.
There are still journalists working who commit to this ideal, no doubt, but their share of collective voice is in the minority. The voices who have been elevated to headline status, whether by newspaper editors or social media algorithms, have it not because of their deep expertise on a subject, as was the case with Krugman, but because they think and write sexy.
The trouble is, sexy is subjective. So to be sexy, writers need to contort their words to their audiences’ lustful gaze in the same way that women have for decades felt the pressure to contort their physical and behavioral selves to meet the expectations of the male gaze.
The other trouble is that “facts” can be just as slippery. What we know is always premised on how we observe and experience the world, the mental models and values we bring to how we interpret those experiences and data, and deeper biases and undertandings about the nature and knowability of the world.
I’m not saying that facts are whatever we want them to be, but that the world isn’t fully knowable by individual methods or minds. Knowledge is a collective enterprise. If someone’s coming to you claiming a monopoly on the truth, they’re trying to take you for a ride.
Because we live in a time of rapid change, frequent conflict, and myriad unfolding social and ecological challenges, many of us share a sense of discomfort or skepticism regarding “common knowledge” and “conventional wisdom”. We feel vulnerable. We feel shamed, perhaps. And information that reassures our preexisting biases is, well, comforting. Its not us its them.
You know what isn’t sexy? Stories that make us feel personally uncomfortable. Stories that make us question the rightness of things about ourselves that we have long held to be true. Stories that force us to be better. Improving requires that we first acknowledge we have room for improvement.
So even when writers do bring up challenging issues, everything has to have “both sides” — which is another way of saying giving quarter and comfort for your readers to be wrong, to reject the information in front of them, to deny the harms being to others.
A world where everything has multiple equally valid sides is a world devoid of morality.
There is no denying that this is the current state of our media ecosystem; neither is there any denying that the both-sidesing of issues related to people’s rights and health and well-being has been an effective strategy for eroding people’s rights and health and well-being. The New York Times is complicit in the wave of anti-trans discourse and legislations we are presently experiencing, to offer just one particularly pernicious example.
Thinking with community
I think the solution here is that we need to foster a completely different model for how we share, collaborate, draw on, and grow expertise.
A handful of particularly sexy thinkers have managed to grab most of the oxygen in the room. You know them — the Ross Douthats and Nate Silvers and Noah Smiths — personalities who have decided that their broad ability with rhetoric, wheter written or on a talk show, together with their technical expertise in a specific area or subject, qualifies them to comment on any and every social issue and current event, from climate change to reiligion, the war in Ukraine to reproductive biology. And they’ve been enormously rewarded for it as of late, with clicks, followers, bylines, and lucritive media contracts.
Shortly after the 2024 election, I saw a lot of people saying that the left needs its own Joe Rogan, after the former reality TV host and mixed martial arts fan appeared to play a role in swaying voters toward supporting the re-election of Donald Trump.
I think this is wrong. The whole premise of a cult of personality is capturing people’s attention and fealty by convincing them they can trust you, no matter the subject, over everyone else. They do this by developing a style of argument that you get familiar with, and then they deploy that packaging over and over again. Readers shift from trusting the content to trusting the voice, often I imagine without realizing it.
You don’t counter this with an alternative cult of personality. You counter it by building more inclusive ways for people to get and share their knowledge: networks of people sharing knowledge and experience built on relationships that are more familiar, more trustworthy, and more accessible. It cannot be one-way. Dialogue is better than lecture. And you have to normalize not just admitting when you are wrong, but taking joy in it so we can learn from it.
Learning > Certainty. Learning is a feature, maybe the singlemost defining feature, of life.
Ethically engaging any issue of public import, from food to religion to gender, requires a proactive attention to multiple perspectives and ways of knowing.
It also requires admitting when you are wrong. I think a lot of the folks occupying the media thought leadership space worry that if they admit they were wrong on something, then their audience will forever question them on future matters. My experience in the classroom has shown me otherwise. You want people to question you, otherwise they won’t be open to being questioned themselves. I tried to run all of my classes as conversations with a lot of give and take. I regularly couched what I had to say with my uncertainty and bias and made a point to acknowledge when someone pushed me in a way I appreciated. And I went out of my way to do this in the first few class meetings. These moments always improved rapport and noticeably brought us together as a group.
These habits have another benefit: they are catchy. By modeling them I was encouraging students totake them up. By mid-semester our classes, which invariably comrpised a group of people who had more or less just met one another, were able to talk about serious and sometimes divisive issues with grace and respect.
Knowledge and expertise are a process. We are always learning. And, notice the “we” in that sentence. Knowledge and expertise are a group process not an individual one.
Last week I put out a plea to scientists to hold fast and fight back in this deeply troubling political moment. To conclude with something concrete, I want to return to those suggestions.
We already have a really excellent example of collective action towards the creation of an open learning commons: Wikipedia. It is an unmatched example of the people creating novel democratic social norms and processes for empowering dialogue and discussion of knowledge. At its core is a truly radical sentiment: all knowledge is revisionary, therefore all knowledge is a candidate for revision. There is no “end state” for an article on Wikipedia, a point that once reached articles are permanently locked from changing.
Fascists hate Wikipedia both because of what it achieves and what it represents.
Wikipedia isn’t enough. I also think we need to create smaller, more community-scaled networks for learning and expertise. Perhaps around issues in general, like food, or maybe about issues as they play out for us in our neighborhoods and cities.
The way we do this is by being generous thinkers— rather than trying to establish ourselves as the next Big Thinker On All That Matters™, we should defer to those we trust, direct people to them, and in so doing become trusted as brokers and connection makers.
Google published really interesting research in 2023 on how Gen Z engages with online information. What’s most noteworthy I think is how these younger thinkers are increasingly learning through social dialogue, intermeshing their experiences, family conversations, and online exposures to information. It also shows just how powerful algorithms can be in shaping those exposures.
What I think we need more than anything right now is to deploy innovative new ways to connect, build trust, and share our own knowledge and experience without the bottleneck of messengers with facile takes or algorithms with spurious motivations.
We don’t need a new algorithm. We need to be the new algorithm.
Bluesky offers an imperfect example. Anyone with an account on the new distributed social platform can build curated lists and starter packs that other users can follow. There are starter packs of journalists, starter packs of foodies, and so on. I trust Katharine Hayhoe, for example, on all things related to climate science (disclosure: I also work for Katherine). And because I trust her, I trust the people she puts on her curated lists of people posting on climate. She has something like 20 climate related starter packs, for climate science, climate finance, climate activism, climate activism in Texas, even a list of “climate Katharines.”
These nascent social connections are the building blocks of the new algorithms, algorithms that adapt and learn based on our collective input rather than based on their success at generating advertising revenue.
Learning how to do and share science in this kind of enriched hybrid social environment and in the context of great geopolitical change and unrest is, I think, among the most important challenges facing society right now. Currently, a handful of people and platforms have gamed the system such that their voices take up most of the space. But we can take that voice back.
It’s a full time job to be a Ross Douthat or a Joe Rogan (what a world to even write such a sentence). For the rest of us, we have to contend with the myriad responsibilities and commitments of our jobs and lives. Individually, we can’t compete with people who are paid to do this 24/7, but if we work collectively, they won’t be able to keep up with us.
