Twitter's content warning algo, equity audits, politicians trading stocks, Fed's next move, mRNA & more

The episode opens with an extended comic bit about the hosts' sweaters. One jokes that his is Tom Ford with buttons made from endangered rhino horn; another names an Italian cashmere brand and turns it into a mock advertisement; a third describes calf-leather with shark-fin buttons. A host then recommends a roughly $150 jacket from a maker called Marine Layer as the down-to-earth option. The riffing is purely for laughs and sets the show's self-mocking tone.

The banter loops in a running joke about "Sweater Karen," a listener who supposedly commented on the hosts' clothing, and a story about a host's wife buying him several sweaters for a mountain holiday for about $3,000, which he says he was not comfortable spending. One host then does the formal intro, noting the show keeps breaking into the top podcasts "week after week" thanks to the audience, and introduces the others with the usual nicknames.

One host explains he skipped two parties to stay home, which leads into his New York stories. The first is a helicopter trip back from a golf outing at Deepdale with what he calls a well-known political figure. He says he has been afraid of helicopters since a landing about eight years earlier, when a skid failed to settle on the East Side heliport pad by the East River and the pilot had to lift off and set down again. He describes it as the most scared he has ever been.

The second story is about a late dinner at Carbone with Draymond, Kevin Durant and friends after a game, told as a tale of broken willpower: he had already eaten a work dinner and brushed his teeth to avoid a second meal, then gave in to wine, bread and pasta and stumbled home at 3 a.m. before an early meeting. A separate thread teases another host for losing a hard-to-get Carbone reservation tied to the Bird IPO dinner. None of this is analysis; it is extended banter before the show turns to news.

One week into Twitter's new CEO, the hosts note that a tweet by Sacks about investors wasting time debating interest rates, inflation and tax policy was flagged with a gentle label: some conversations can get heavy, and there is a human behind the screen. They debate whether the flag is machine-generated or applied by an editor, and puzzle over what in such a mild tweet could have triggered it. One suggests the label is aimed at the replies rather than the tweet itself, to tamp down insults aimed at the author.

Sacks reads a political motive into it, calling it the left going after the right, but another host pushes back that he has not seen left-leaning accounts flagged and reads it as a blanket attempt to soften the tone of replies. The group agrees the key open question is whether the system is algorithmic or editorial: if editorial, people with an axe to grind could flag selectively; if algorithmic, it needs tuning. They call for the transparency about ranking algorithms that Jack Dorsey promised in congressional hearings and never delivered.

The hosts argue the labels should at least explain themselves, for example stating that an algorithm flagged heated language or that users reported the thread. One concedes that any centralized network will involve some moderation, and that Twitter's or Facebook's platform is ultimately their own marketplace of ideas. The segment turns comic again as flags appear on other hosts' tweets and profiles, prompting jokes about needing a "meme team" and paying per viral post. The substance here is thin; the real point is the demand for transparency.