Tech layoffs surge, big tech freezes hiring, optimizing for profits, election preview & more

The episode opens with banter about what one host is wearing: an Uber-Moncler crossover hat, bought along with a Moncler shirt. The hosts riff on imaginary Moncler products, including an Apple watch, a commemorative mug and a neck tattoo, after one speaker says Uber had a big week.

The joke then turns to sponsorship: one speaker hints at a $100,000 deal and asks whether their agreement covers logo placement. Another replies that anyone willing to sponsor him would cover every square inch of his clothing with ads, like a race-car driver. The hat, the speaker says, cost $55 plus $17 shipping from France via Etsy. The opener cuts off mid-sentence at "And it said We open", so the source breaks off here.

Asked how their week went, one speaker says he has been busy working and trying to be helpful. He then addresses inbound questions about whether he is leaving his venture firm for a role at Twitter. His answer is no: he and another host are pitching in while Elon sets up a permanent team.

In this account, Elon is the CEO and the decider, and the two hosts are part-time helpers assisting a friend. One speaker says the media has blown this up into something much bigger than it is, and that he is still doing his day job: podcasting and investing in about a hundred companies a year.

One host raises an issue he says is already public from a tweet that morning: an advertiser boycott, following reports of a flood of racist tweets. His account is that these reports are not true as framed. He says that within hours of Elon taking over on Friday, people from a message board, in what he calls a "4 channel attack", built bots that posted huge volumes of spam messages containing racist words. He says the trust and safety lead, Yoel, met with him and another host, and Elon directed him to shut it down.

The speaker argues this was unfair to Twitter because the posts came from spam or bot accounts with zero followers, or followers that were other bots, so no real user's feed was degraded. He says it was shut down promptly. These are his claims, and the transcript offers no independent check of the details or the figures.

He then describes how, in his telling, a boycott gets built: activist groups monitor the firehose, publish a report that racist tweets rose 500% since Elon took over, feed it to news outlets, and someone passes the coverage to advertisers. He adds that Elon had not yet changed any content moderation policy, and that the incident was manufactured by people not acting in good faith to hurt the company. Again, this is his characterization.

A speaker recommends episode 333 of the podcast the transcript renders as "The Lux Freeman podcast", an interview with a guest it renders as "Andre Carpathi", described as one of the great minds on AI and machine learning. The question discussed there, as he relays it, is what the next generation of bots will look like.

His own takeaway is that this is hard for every platform, not just Twitter. Tools can now generate realistic, unique human images and high-quality text through systems like GPT-3, pushing the boundaries of a low-level Turing test. Over time, he says, detecting and blocking spam and coordinated attacks will demand much more sophistication.

Another speaker notes the guest used to head Autopilot vision at Tesla. A speaker then says he has no doubt Elon will do a much better job stopping bots once he has time to do it, given his team of AI engineers and sharper focus. That is an expectation, not something shown in the transcript.

One speaker likes the idea of micropayments or subscriptions to third-party content providers inside Twitter. Much of his news arrives through the app, and clicking through often hits a paywall or other friction. Being able to buy a single article, he says, would be a super feature.

He also wants Twitter to experiment with letting publishers post without a byline, in the style of The Economist, so credit goes to the masthead rather than the individual. He thinks that could change how journalists cover the news. Another speaker adds that if publishers are paid enough, a platform could demand that today, with items simply labeled by outlet such as the New York Times.

A speaker says The Economist is polarizing in journalism for this reason, and that many journalists would prefer to lose their bylines. His argument is that even good-faith reporters on some topics can face harassment and threats once their name is attached, so many writers might opt in. Another speaker says they should, preferring two clear ends of a spectrum to the "gross middle" that exists now.

On the one end, he places the New York Times, The Washington Post and The Economist with no bylines or attribution to individual reporters. On the other, writers who want a brand built on their own name should start a Substack. He argues that separating the two would make news more likely to be truthful news rather than well-disguised opinion.

A counterpoint follows: a no-byline publication gets little time to build trust, and readers may suspect the byline was removed to serve an agenda. The Economist, he says, built that trust over decades. Substack, by contrast, shows that someone with no reputation can build a business from scratch with good content. His conclusion is that the Times should focus on being the Times and Substack on individuals, and that cleaning up the middle would help everyone. The segment closes with a speaker saying he is glad the others are pitching in and would like to use Twitter more often.

A speaker asks to discuss the episode in which two men posed as fired Twitter employees and were reported on by journalists. In his telling, reporters parroted it because it fed their narrative, with no back-checking and no reporting. He says several outlets apologized, including a CNBC reporter whom the transcript renders as "Dear Grabosa".

As described, the two men walked out of Twitter's offices on Monday carrying a box and said they had just been fired. The names the transcript gives are Rahul Ligma and Mike Johnson, which combine into a crude joke. One speaker says his own pre-teen children got the joke immediately, which he treats as a telling sign for the news industry.

The hosts' reading is that journalists did not catch it because they did not want to, since the story fit their narrative, and that they do not fact-check what fits. One speaker calls the moment poignant, given the importance of open or citizen journalism and the trust placed in journalistic voices, and says the men were able to con reporters right outside Twitter's offices into a sensational story.

To show the pattern, a speaker recalls a story he thinks first ran in Rolling Stone and was then amplified by Rachel Maddow: that in Oklahoma City, MAGA Republicans were taking horse paste as a COVID therapy and crowding emergency rooms so badly that heart attack victims were turned away. In his account it turned out to be a hoax that the media reported because it fit its preconceptions. That is his recollection, and the transcript does not verify it.

Another speaker adds a separate factor: live coverage. In real-time reporting, people call in from the scene and the outlet wants to carry it, so the hedge, that something has not been confirmed and two employees are merely claiming it, matters. He says real-time coverage is fine if it is presented with that caution.

The exchange then gets heated. Someone jokes about a picture captioned as Ligma Johnson "the copy editor", another speaker calls that lazy and asks why he is covering for these people, and the first insists he is only explaining. His explanation is that live news has a different standard and that fact-checking has been stripped from many publications amid a race for clicks.