Ask for the Definition First: David Pakman on Labels, Trump, and a Tweet He Should Have Left Up

On Monday, March 27, 2023, at 3:42 p.m., David Pakman posted a snarky tweet after a mass shooting at a Christian school in Nashville. The point, he tells Lex Fridman, was that thoughts and prayers neither prevent shootings nor do much afterward, and that it is odd to blame a lack of prayer when the shooting happened at a school that prays. He glanced once that evening: about 2,000 likes, a few replies saying he had missed the mark, and then he went on with his day. It was one tweet among some 20,000.

His baby daughter woke him at 6 a.m. While making breakfast he saw verified accounts quote-tweeting the post and demanding a retraction. Around noon his father called: a hundred direct messages, ranging from “you should have aborted your son” to “we’re gonna find all of you.” Emails arrived about Pakman’s Jewish faith. He deleted the tweet, hoping that would stop people from hunting down relatives. It did not. Screenshots were already everywhere, and over the next 48 hours the pile-on moved through Candace Owens, Fox News, and Newsmax, peaking with Donald Trump Jr.

He now calls the deletion the move he regrets. Once he told his audience what had happened, thousands wrote that only someone determined to read him in the worst faith would think he was blaming murdered children, and that he should have left the tweet up so the tide could turn. Asked whether the post actually represents what he believes, he does not hedge. It “absolutely” captures his disgust with a politics that will not touch guns and would rather point at mental health or more prayer in schools. It is not, he adds, the sarcasm he would use if he had an hour instead of a character limit, and he knows the wording was “unnecessarily provocative.”

Fridman opens by laying a pile of labels on the table: liberal, Democrat, left-wing, leftist, progressive, socialist, communist, Marxist, far left, center left. Pakman splits the problem in two. Academically, the terms have different meanings and can be defined as a precursor to talking about policy. In practice they are often tags, used to hand the listener an image of a person rather than a claim about what that person would do. Marxist, he says, is currently popular on the right as an attack on Democrats, even though actual Marxists in positions of power in the United States are vanishingly rare.

The more interesting split, for him, is liberal versus progressive. Ten years earlier, liberal often meant what people now mean by progressive. More recently the socialist-leaning left started using liberal to mean Joe Biden: someone not left enough. About eight years before this conversation Pakman told his audience he would start calling himself progressive, because liberal had been turned into a smear and had also shrunk toward center-left politics. Inside his own media world, one attack on him is now that he is “just a Lib,” meaning insufficiently left, progressive, or socialist. Callers use the words casually and almost never define them, so they more often stifle conversation than encourage it.

He refuses to treat progressive as a synonym for democratic socialist, and he pauses on a linguistic trap that trips a lot of American arguments. Social democracy, in his usage, is a highly regulated form of capitalism of the sort associated with Northern Europe and Denmark. Democratic socialism is a form of socialism that would socialize ownership of the means of production. Moving from one to the other, he says, is a fundamental shift in how society operates. The United States has already socialized the military and, in a limited sense, some health care through the VA and Medicaid. Social democracy would socialize a few more things inside a capitalist country. He is not a democratic socialist, and he does not pretend to be the best advocate for how that system would work.

Fridman brings up Pakman’s tense appearance on Patrick Bet-David’s podcast and says he thought both men did well, with Bet-David playing a stubborn devil’s advocate. Pakman got the full range of mail, from “best thing you’ve ever done in public” to “your mother should have aborted you.” The feedback that interested him most was people asking whether it had been unbearably awkward afterward. For him, that intensity is a comfort zone: he disagrees, and then it is over, with no grudge. He does not walk into every studio the same way. A sit-down with Jesse Lee Peterson in Los Angeles is a different event from a talk with Bet-David. He stopped going on Michael Knowles’s show because of the volume of threats afterward, even though he found Knowles unusually clear about his own views, which made the conversation itself easier.

Fridman asks whether Twitter sarcasm can turn a thoughtful commentator into a partisan hack. Pakman does not offer a grand theory. Across disciplines, he says, it is not the best way to solve complex problems, and he does not think Twitter has been an unmitigated good for society. He is on it anyway, because it is one of the pipes he uses to get a message out, good and bad included. He also has to shape content to the platform, the same way a YouTube segment is not a TikTok. Fridman worries that a machine of Twitter and journalism is addicted to the drama of division, and that mocking ridiculous people on the other side is a drug. Pakman agrees.

That agreement is how he describes his show. He gets two kinds of email: do more sensational pop-culture content and you will get bigger, or do a deep dive on Calvin Coolidge, which he is sure nobody would watch. In a given hour he will mix a relatively serious look at the federal budget, some political “tomfoolery” at the post office, and a clip of Trump inventing words at a rally. He uses the old food pyramid as a confession. Junk food is at the top, and “the vast majority of what I do is the top of that pyramid.” He tells viewers he does not consume much of the type of content he produces. The base should be critical thinking and epistemology; then history, economics, and philosophy; then current events from primary reporting such as the Associated Press; and only after that the commentary he makes, if they find him thoughtful enough to earn a place there.

Fridman asks about audience capture: whether an audience recruited by mocking Republicans will train the host. Pakman worries about it and is relatively comfortable that it has not taken over, because he knows which topics upset his core viewers and he still touches them. Israel and Palestine, vaccines, and some foreign policy will do it. So will his lack of enthusiasm for Marianne Williamson and Robert F. Kennedy Jr. as challengers to Biden, which is not about love of Biden but about whether they are serious. The people who contacted his advertisers after the Nashville tweet, some of whom got him dropped, were not his audience. They had been, in his word, weaponized. Followers of Candace Owens, shock value aside, are not the people who watch his show.

If the 2024 election were held the day of the taping, with Biden and Trump both announced, Pakman thinks Biden wins. He starts from 2020 and asks what has since moved voters toward Trump; he cannot make that case. Florida has moved more Republican, but Trump already won it. Arizona and Wisconsin, in his reading, have moved left. He does think Biden is a worse candidate than in 2020, and he admits that sounds ageist: the man is older, Democratic voters want younger options, and in January 2025 he would break his own record as the oldest president ever inaugurated. The other side of the ledger is a term’s worth of work, including the Inflation Reduction Act, COVID stimulus, and a pile of smaller healthcare and student-lending tweaks. Macroeconomically Pakman does not give Biden, or Trump before him, too much credit, but inflation had started to come down and the stock market was steady, which is historically a decent environment for an incumbent.

Fridman offers a personality comparison. Trump’s flaw is chaos, to the point of disrespect for groups, ideas, nations, and leaders. Biden’s is that, perhaps because of age, he is not mentally present in the way Barack Obama was. Pakman accepts both sketches and says the meat is how you weigh them. “Biden’s gonna get us into World War III” did not seem to be happening. He does not see Biden’s deficits as dangerous to America’s standing with traditional Western allies in the way Trump’s unhinged personality, combined with a lack of knowledge about most issues, is a threat. So if those two are the choices, Biden is his. He would rather see someone else on the Democratic side, and if he knew the next president would be a Republican he could name better options than Trump.

The contrast with earlier cycles is the point. Obama versus Mitt Romney in 2012 felt enormous at the time; looking back, Pakman would still disagree with Romney on tax rates and LGBT issues, but the difference now looks comparatively benign next to four years of Trump. John McCain versus Obama also felt drastic, and Pakman locates a hinge: McCain’s choice of Sarah Palin as running mate “opened the door to the sort of cartoonish stuff” that later included Trump and Marjorie Taylor Greene and has made the party, in his phrase, a bit of a joke. You can argue that taxes are too high. “Jewish space lasers,” he says, is another matter.

Pakman places himself as a progressive social democrat, which he stresses is still a capitalist: he owns a business and pays the taxes the law requires, “not a penny more.” He puts himself to the left of Biden. On Alexandria Ocasio-Cortez, he talks less about a policy list than about durability. Unless she hits a major political problem, not a personal scandal, she looks to him like someone with staying power in elected politics. He has no idea whether she wants the presidency, the Senate, or a governorship. Policy aside, she combines charisma, likability to some, and a personality that angers people who already dislike her in a way that only fuels her presence, a pattern he also sees in Trump.

Fridman asks whether the future of campaigns is whoever can most skillfully infuriate the other side. Pakman says that is one aspect, and that communicating policy still matters. Then he offers a rare bit of praise from a Trump critic. Trump often says strange things that, transcribed, barely parse as language. In the moment, the way he relates to adversaries on stage is very good, because he knows how it will look to the audience. The word salad matters less than the immediate response. That is one reason Pakman finds it hard to believe Ron DeSantis can beat Trump in a national primary. Being good in clips, often shared out of context, and knowing how to use social media, which means something different every cycle, now outweigh a deep dive on the deficit. He thinks that shift is bad. He also thinks it is real.

Bernie Sanders, who calls himself a democratic socialist, is the hard case. Pakman has listened to him for years, back to the Thom Hartmann show, and by the definitions used in this conversation he does not think Bernie has been a democratic socialist in his public-facing campaigns. He has heard Sanders talk about social programs and their value in American history, not about socialism as ownership of production. There may be 1970s material that is the real thing; Pakman was not there. He puts Biden at center left, himself to Biden’s left, and maybe just inside AOC and Bernie, or very similar to Bernie, with more flexibility on implementation. Medicare for All, expanding the existing program by dropping the age limit and paying through taxes, is one path. He is open to other models, including Singapore’s and Germany’s, if they get everyone good, affordable care. Whether that puts him to Bernie’s right he does not know. Fridman calls it flexibility versus dogmatism rather than left versus right, and Pakman does not disagree.

Asked for Trump’s strengths, Pakman says they are mostly superficial and about presentation. A rich New Yorker convinced people he had spent his life trying to be isolated from that he had their interests in mind, understood why they were not doing well in 2016, and had solutions. Pakman does not think Trump liked those people, understood the causes, or offered serious fixes: Jared Kushner would solve Israel and Palestine in year one, Obamacare would be replaced in 2017, Mexico would pay for a wall by the end of the first term. Callers to his show believed the wall. The 2017 healthcare proposal, he says, would have left about 24 million people without coverage and went nowhere; in August 2020 Trump promised a plan in two weeks, and by 2023 it had not arrived. Fridman pushes back with DeepMind, OpenAI, and SpaceX: sometimes you need someone naive enough to rush in. Pakman agrees, then notes that all of those successes are from technology, not politics.

On China he finds a positive case hard to make. In a recent Fox interview with Mark Levin, Trump suggested that tariffs had cost China so much that Beijing released COVID as punishment. Pakman says American companies pay those tariffs, that Trump still does not seem to know this, and that reports put the job cost of the China tariffs at about a quarter of a million American jobs. As a revenge plot the virus is an odd instrument: of 6.8 million deaths worldwide at the time they were speaking, 5.7 million were outside the United States. He grants that Trump’s willingness to take risks with a superpower is an argument some people make. He wants tangible positive outcomes and finds them tough to identify. It is true, he adds, that no new wars started under Trump, and he likes that. He does not buy the slam-dunk claim that Putin stayed out of Ukraine only because he feared Trump.

On COVID, the con is brutal: “He’d be President right now if he had handled it differently.” Early on Trump treated the virus as something that would go away, fifteen cases becoming zero, the country open by Easter 2020. Pakman has made the same point to Joe Rogan and Patrick Bet-David: if Trump had asked for help across party lines, even sold MAGA masks, he thinks the man wins reelection. Fridman notes that the administration fast-tracked vaccines. Pakman says Trump touted that until his own crowds turned, then drew a line — he made the vaccines, nobody has to take them — and got booed at his own rallies. The nine-month timeline, in Pakman’s telling, rode decades of mRNA work; what Trump did was pre-purchase supply so manufacturers could scale, which any president in that seat would have done, and he did not get in the way. Both men agree the public communication was a disaster. Pakman would have preferred officials saying they did not yet know whether masks did X or Y, and that supply had to be reserved for front-line workers. He also thinks that if Trump had still been president when doses went out, as many Democrats would have gotten vaccinated and far more Republicans would have followed a political leader rather than the science.

Biden’s weaknesses, Pakman says, include the ones Fridman already named. He is not seen as high energy. He is not the Biden who ran circles around Paul Ryan in a 2012 debate. He does not inspire the way Obama inspired people coming into politics, and on policy he is not that interested in what younger voters want, including cannabis reform and student loans. The upside is foreign policy, diplomacy, and high-level negotiation: it is tough to find someone with more experience, even if critics say that was the old Biden or that decades in Washington are themselves a sign of rot. Pakman’s sense from people he talks to is that Biden is involved in the highest-level geopolitical and big domestic economic questions and is not plugged into a lot of lower-level detail. Trump, he notes, would dive into the micro when it was advantageous — a Sharpie on a weather map, paper towels in Puerto Rico — and otherwise wanted the wall simply built.

Fridman is troubled by how viciously Americans talk about whoever holds the office, including the reply pile under Barack Obama. Pakman describes an approval trap: if a president ever did one thing you disliked, many people will say they disapprove from that point on, so time in office itself drags the number down. Unwinding that would mean changing the election system, including first-past-the-post and the Electoral College. He also thinks Americans who talk politics have become uncomfortable with disagreement in ordinary life. He is from Argentina, where heated argument at a gathering is common and then everyone goes back to cooking the beef. Earlier that day a stranger had said “RFK all the way baby,” then grown awkward when they ended up standing together waiting for sandwiches. Online, he is sure, most of the people who send threats would, in person, discover they share a diagnosis of the problem and differ on four issues.

He does not regret mockery as a tool. It is a calculated part of a package that also includes simplifying complicated ideas and choosing undercovered stories, and he knows his show is not “the audio visual version of a peer reviewed” trial. Candace Owens had blocked him for years and unblocked him only to tweet about the Nashville post; she, Donald Trump Jr., and Ben Shapiro are welcome on his show. On YouTube metrics he treats view counts as a tool, not an obsession, and he thinks about a week of shows rather than one clip. A segment with ten ideas for reducing gun violence will not get half a million views on day one, and he does it anyway. The metric he cannot control is the election cycle itself, which can quadruple an audience in October and collapse it after Inauguration Day.

Pakman has met Robert F. Kennedy Jr., once had dinner with him, and shares friends, which makes the verdict awkward. He sees the campaign as a chaos candidacy to raise awareness and perhaps money for a book or an anti-vaccine organization, with reporting that Steve Bannon liked the idea of Kennedy running as a Democrat. He finds the anti-vaccine work uninspiring and notes that it long predates COVID shots. Marianne Williamson is another sore spot with his audience. He likes the idea of presidents from outside the lawyer-and-business pipeline, including education, science, and nonprofits, but Williamson’s “woo-woo” background puts him off. On Russell Brand’s show she said there is no such thing as clinical depression, only someone in a clinic telling you that you have it; when asked later, she said she had been trying to impress Brand. Pakman is not sure which bothers him more. He agrees with her on climate and healthcare access, and he does not want an uncontested Biden primary. He would like debates above some polling threshold. He does not expect the DNC to organize them.

On the Republican side he thinks DeSantis, in the weeks before the taping, had begun to implode before formally starting. Asked about Trump’s attacks during a trip to Asia, DeSantis went into what Pakman calls a bobblehead smile and got angry at the reporter; asked about Guantanamo, he attacked the journalist. If you cannot handle that, Pakman says, you cannot share a debate stage with Trump. Big donors looking for a non-Trump basket were already saying the governor did not have it. DeSantis was polling in the twenties without announcing; Trump was in the low fifties, which means DeSantis would need more than the entire non-Trump vote. Nikki Haley looked dead on arrival; Vivek Ramaswamy was also running. Trump actually does better, Pakman says, the more people pile in.

He sorts Republican voters into four groups. The old three are pro-business low-tax types in the Mitt Romney mold, libertarian-leaning voters more interested in freedom than in social conservatism, and religious conservatives around figures such as Josh Hawley and Ted Cruz. Trump introduced a fourth: people who had not paid much attention to politics, followed celebrity, had grievances without a scapegoat, and were culturally right-leaning without calling it Republicanism. In 2016 the mobilizing issues were real economic pain in places such as Ohio and Indiana, plus scapegoats at the border and in China, not yet the later culture war. Populism, he stresses, is a rhetoric, not a policy kit. Bernie and Tucker Carlson can both talk about the middle class and then propose opposite remedies. Hillary Clinton’s unique heat, as he lists it, came from time in public life, Trump’s “crooked” brand, the email story and James Comey’s late public interventions, and a likability problem even among people who voted for her, as he did.

Why do thin conspiracy theories stick? Pakman thinks many people find random, unpredictable disaster harder to live with than the idea that someone planned it and that it could have been stopped. He has no problem with suspecting a public figure is holding something back. The problem is losing a shared method for deciding what would count as true, so that both bad evidence and missing evidence get recruited for the same theory. He compares it to arguments between atheists and religious believers about whether desire that something be true is a way of knowing.

Was January 6th a big deal? Compared with the Civil War, Pakman says, less of one. The larger question is what the six hours, the run-up, and the aftermath did to what citizens treat as on the table after a lost election. In the summer of 2020 Trump was already saying they were going to cheat, even though polling made a Biden win unsurprising. Vote-counting schedules by state were known and followed. At 2 a.m. Trump came out and said he had won. From there came millions in donations for undefined audits, lawsuits that went nowhere, a proposal to send alternate electors from states Biden had won, pressure on Mike Pence, and then the Capitol. By December 1, giving three weeks to look, Pakman says there was no reasonable case that Trump had won. That sequence is, for him, the height of no longer attaching politics to a verifiable factual reality, and of being willing to commit property and violent crimes to get something other than democracy.

Fridman argues that tanks in the streets from a more effective authoritarian would have been worse, and that a few protesters in a questionably protected building might still be protest. Pakman agrees that a military move would have been worse and does not want that comparison to minimize what the protesters intended, which was that Trump remain president. He does not believe there is reporting that Trump actually tried to use the military. The playbook, though, is already in use: Kari Lake lost the 2022 Arizona governor’s race in the ordinary sense that her opponent got more votes, and she still insists she won. If Biden loses in November 2024, Pakman would bet money that he concedes and leaves in an orderly fashion. If there were real evidence of a hacked election he would want that claim made. He does not expect it on the pattern of recent cycles.

On Hunter Biden’s laptop, Pakman’s rule is simple: if there is reason to believe Hunter, Joe, or anyone else in that orbit committed a crime, investigate, charge, and try them. After years of talk, including the claim that “the big guy” got 10 percent, he says he has seen zero actual pieces of evidence of a crime involving Joe Biden. What the story mostly looks like to him is nude images released without consent, illegal in most states and a violation of Twitter’s rules. Tucker Carlson saying he mailed himself a USB stick of the supposed mother lode and lost it in the mail is, in this telling, almost satire. He is against censorship short of illegal content and respects a company’s terms of service so long as they are not themselves illegal. He finds dubious the idea that seeing Hunter Biden’s genitals would have flipped the election. If the articles had not contained those images, he would have wanted the links allowed.

Tucker Carlson’s firing from Fox, Pakman says, was not a political conversion at the network. It was risk mitigation. Carlson’s show was a prominent node in claims that helped produce a $787.5 million Dominion settlement; Smartmatic still had a similarly sized suit; a former staffer had a discrimination case that looked to him as if it had legs; and Ray Epps, featured on *60 Minutes*, was another lawsuit waiting to happen. Fox is also less dependent on ads than CNN or MSNBC because it collects a fee from every cable subscriber who has the channel, whether they watch or not. Pakman thinks Carlson’s most profitable path is to go independent rather than to Newsmax.

The meme that both sides are the same is, in Pakman’s mouth, a conversation killer. Some people use it as what he calls enlightened centrism, a pejorative: I am above the fray, two sides of the same coin. Others use it when their own side has stepped in it, to admit a wrong without admitting it is unique. It also, without anyone needing to say “don’t vote,” suppresses turnout, and he wants turnout high. He does not pretend to be above his own bias. In 2023, if the American right got everything it wanted, he thinks the result would be horrifying. If the left got everything it wanted, the country would have to figure out how to pay for some programs, but the goals are mostly noble and, in his view, more aligned with individuals determining how they live.

Fridman turns to the cultural charge: censorship, speech codes, reallocation away from merit, the “woke mind virus.” Pakman cites polling that did not use the word woke and still found the “hasn’t gone far enough” position on issues such as the trans community more popular than “gone too far,” often in the high fifties against the low forties. That is one reason he thinks DeSantis’s anti-woke agenda may be a political misstep. Four years earlier, woke had meant becoming aware of problems bigger than any one person can fix. Now it is a pejorative for outrageously left-wing, maybe with Marxist undertones. He has lived a small version of the excess: as a Jewish Argentine immigrant he has been told he is not oppressed enough to speak, and he dislikes the Spanish-language push to replace gendered words such as *amigos* with *amiges*, a change nobody in his family uses. He tells people on the left they are hurting their own movement. He does not think it is as big as critics believe, and he does not think it is an existential threat. Under Biden, whom he sees as center left, that culture war has played almost no role in more than two years of governing.

Trans women in a short list of sports is, he says, a real and complicated question he would take to leagues with experience. Pretending that the vanguard of left-wing politics is forcing trans women into sports is, in his view, the right reacting as if a niche issue were the main event. He knows people leaving academia because of campus climate, and he rejects a fantasy of perfect meritocracy. What he wants is roughly equal opportunity, a bell curve of outcomes, and the chance to pursue a path without suppression. Fridman worries that universities now listen to the most outraged 19-year-old and would not tolerate a young Noam Chomsky. Pakman sympathizes and asks for examples: of five he is sent, one is often a real case and four are a complaint that never threatened tenure.

The four-year degree, Fridman thinks, will not look the same in 20 years because of ChatGPT. Pakman jokes that his entire show is already written by it, then notes the model’s knowledge cutoff. He has a 10-month-old daughter and a 529 account, and he can do the tuition math: at current rates of increase, the schools he attended could run $200,000 a year by the time she is 18. What he wants from any new educational core is critical thinking, epistemology, philosophy, and media literacy, skills he finds missing in many people he hears from. Fridman says ChatGPT already helps him think by laying out hypotheses on questions such as a COVID lab leak. Pakman asks whether it also teaches you how to weigh the paragraphs; Fridman says you can literally ask it. On American-style libertarianism, Pakman’s objection is sociological: past about 150 people, some decisions get centralized, specialization appears, and the problems look like government whether you assign them to a firm or a state. On political-compass tests he lands as a left libertarian, anti-authoritarian when that option makes sense.

Pakman has not paid eight dollars for a Twitter check. The old verified replies were useful; the new badge, given how few people have bought it, is not. Cults of personality mean nothing to him. He invested in Tesla in 2015, later sold the shares, and is on his second Tesla, unsure he will get a third because the rest of the industry is catching up on range and charging. What Musk did to force battery and electric-vehicle technology forward he calls a “one person wrecking ball” in the best sense. The Twitter purchase he does not really get, especially after Musk tried to exit and was forced to close. The For You feed is, in Pakman’s daily use, mostly right-wing content he does not want. He does not think Musk is a right-winger; interviews still read to him as socially moderate or even somewhat on the left. It does seem that during the Twitter saga Musk ended up closer to voices that give people the opposite impression.

Fridman suggests that vicious attacks from the left push people such as Musk and Joe Rogan rightward on the surface. Pakman’s experience is different in volume if not in heat: when the left attacks him it is almost as vicious as the right, but it is a smaller contingent, which he notes is partly because he is on the left. The “not left enough” bullying is real. He would still sit down with Rogan, stand by every clip he has done criticizing him, and expect it not to be taken personally. The job itself is “completely and totally toxic”: partisan news on the internet with a social media layer. After the last show of the week he tries to tune out until Monday. He gives himself 15 minutes a day on social platforms. When a pile-on like the Nashville tweet hits, the discipline collapses and he finds himself sweating and scrolling. He would rather have fresh views each day than views flattened by what he thinks people want.

Production is a Letterman schedule: five episodes in four days, Monday through Thursday. He picks six to eight stories each morning, records in the early afternoon, and publishes the same day, working from a MediaWiki rather than paper so the team can share notes. Ideas come from subreddits, a private Twitter list of journalists, AP and Reuters, a glance at Drudge and CNN, and Google News. For facts he wants a primary source first; Wikipedia footnotes and ChatGPT are research tools, the latter only with aggressive checking, because it will invent a source. On a breaking legal or medical question he will call a friend who is a lawyer or a doctor. The hard problem of truth, as he now frames it, is getting people to agree on what is a fact and what is an opinion. A useful opening with someone who disagrees is, “How do you think we would figure out X?” If no evidence could change their mind, he is inclined to end the talk.

Some people on the left of his audience call him a Netanyahu shill, even though he has never supported Benjamin Netanyahu. He thinks black-and-white characterizations of Israel are inaccurate, and that “are you criticizing Israel enough?” has become an American litmus test he will not take. Big picture, he does not expect progress until pre-negotiated terms exist and the parties at the table are acting in good faith. He does not see Likud as a particularly good-faith arbiter of peace, because he thinks the party benefits from the absence of peace and from the threat, and the fact, of violence. He does not see Hamas filling that role for Palestinians either; the Palestinian Authority is a question mark. The most likely, imperfect shape he can imagine looks something like Clinton-era borders, as many settlements as possible gone, land swaps of the kind Yasser Arafat seemed open to in the 1990s, and East Jerusalem as part of a Palestinian state. He goes back and forth on how large a role the United States should play, given that it funds Israel; credibility on the global stage, and a deal in which both sides renounce past claims, matter more than Washington’s desire to be in the room.

On Ukraine he does not pretend to be an expert. As a general observer he does not find it controversial to call the February 2022 invasion an act of aggression by Vladimir Putin. He thinks that if Trump had been reelected, Putin might have reached for other tools of influence rather than a geographic grab, but he does not know that. He has a hard time imagining a face-saving end that is not tragic. What he dislikes on the American right is the reflexive claim that defending Ukraine means supporting neo-Nazis, which he hears as a proxy attack on Biden. Marjorie Taylor Greene’s line that the United States should do nothing, that Volodymyr Zelenskyy is a comedian, and that the mission is to support neo-Nazis is, for him, the issue collapsing into a partisan litmus test. He used to think that contact with intelligent aliens would make human conflicts look small; people in space science told him that after a momentous period it would become normal and the old fights would return. There will always be conflict, he thinks, even if the tools become cyber or biological rather than missiles. He disagrees with Americans who cast the United States as a kindly global police force, and he also disagrees with an older Chomsky-inflected picture of America as a uniquely bad actor; much of it, in his reading, is an outcropping of size and wealth rather than uniquely dark intent.

He reads almost no modern political books or official memoirs, on the theory that he already gets enough of that at work. Narrative nonfiction is the genre he recommends for history and psychology: Neil Postman, Jenny Odell’s book on saving and spending time, and Alfred Lansing’s account of Shackleton’s Endurance voyage, which he finds almost inconceivable. He has read Viktor Frankl more than once. Gardening helped with the toxicity of the job; with a baby and travel, there is no garden this year. Estate planning forced him to think about a YouTube channel that might, grimly, earn more after a spectacular death, and about whether a voice clone could keep the show going. He is more interested in that than in the tweets.

Advice to young people collapses, almost always, to start right away. People who want to do what he does and spend a year thinking about lighting, he has learned, often never start. Immediate data on what works, whether you even like the work, and a cheaper cost of changing direction beat a perfect room. He thinks humanity makes it to 2100 and refuses to name a single biggest threat, because if this is an inflection in technology, 2090’s dangers are hard to know from 2023. Nuclear war, climate, and another pandemic are the clichés; space exploration is, for him, not an excuse to ignore problems on Earth but a way of learning things that can be used here. The night before a trip he always wants to cancel. He never regrets going.