Steve Hilton

The episode opens with the two participants recalling how they met: their young children were digging in the sand together on a beach in Hawaii, two girls on one side and two boys on the other, at roughly the same age. They place the meeting around 2012 and describe a friendship of more than 10 years that predates any talk of a run for governor.

The guest's earlier appearance is then pinned down as episode 509, recorded in 2014, with the hosts noting that about 2,000 episodes have been released since, or 12 years of the show. The guest recalls arriving with a background in English politics and government, business and running restaurants, and teaching at Stanford, yet not knowing what a podcast was. They assumed it was something like a radio show and kept wondering when the next break would come, only to discover the conversation ran three hours.

The guest remembers being struck that the conversation covered sex and was full of swearing, unlike anything they had encountered. The reply was that hardly anyone was listening back then, social media was smaller, and it was possible to talk freely without worrying. Asked whether that worry exists now, the answer is simply not paying attention, described half-jokingly as the key to success.

The conversation turns to criticism online. One side argues that many people are crippled by others' opinions; feedback has real value, but the internet holds too many bad-faith actors who promote the least charitable narratives about any person or subject. The point is made with an analogy:

if you went to a restaurant and you knew that 20% of their food would literally fucking kill you would you ever eat there

The guest traces their own rise in visibility, from private citizen teaching at Stanford, to starting a business, to hosting a show at Fox, to becoming a candidate. They say the abusive comments have grown, citing examples such as being told to go back to England, and admit to a constant urge to reply "good point" that they have never acted on. The response is that sad, angry people are overrepresented online because they have the time to comment, and that in politics much of the engagement does not come from real people at all.

The bot claim rests on the dispute around Elon Musk's purchase of Twitter. Twitter alleged only 5% of accounts were bots, while a former FBI figure working in digital intelligence reportedly ran a forensic analysis and put the figure at probably 80%. The participants are unsure of the purchase date, guessing 2023 or 2024. A lookup read aloud during the show complicates the 80% claim:

while official company filings have historically claimed fake accounts represent under 5% independent estimates and security specialists suggest real figures are higher though an exact 80% total remains heavily debated

The source of that figure is named as a 2022 assessment by cybersecurity expert Dan Woods. The rebuttal on the show grants that the number is debated but argues the article is meticulous. It contrasts this with Twitter's method, described as sampling a random 100 tweets, and notes the company had a motive, since it was being sold for a sum the participants recall, without certainty, as about $40 billion.

The discussion goes on to bot farms: whistleblower footage is said to show rows of Android phones wired to keyboards, where one typed message goes out from hundreds of devices. The participants argue AI now makes such campaigns easier, so anyone could hire a company to push a smear across platforms. They also describe paid human posters, called out by independent journalists for using identical wording, and claim Israel pays people and has been open about wanting to win a digital war. Real people and real independent journalists remain on the platform, they say, alongside "a tremendous amount of just utter horseshit."

The guest describes campaign-era habits: no time for Instagram or other platforms, but X checked for perhaps half an hour in the morning, half an hour midday and half an hour at day's end. X is valued as "the best thing in the world for news," while opinion threads are skipped. The abuse, the guest says, just makes them laugh.

The host recalls that the guest went years without a phone on them and could be reached only through their spouse, an approach that once went badly in San Diego. Phones are framed as a toxic, distracting element that keeps people from being present, and the host admits to inconsistent discipline, including 20-minute sessions of watching disturbing videos. They describe a morning tradition with friend Tom Segura of trading the worst footage they find, such as a failed ATM theft in which a man was dragged under the machine, and they say Instagram is poor at removing execution videos.

The comparison between platforms is specific. X is described as mostly opinion and text, while Instagram also surfaces historical facts and scientific discoveries. Instagram appears to be what people watch on planes, though TikTok may be bigger and skew younger. TikTok is singled out as more addictive because it starts playing content the moment it opens, unlike Instagram, which requires some effort to browse. Both recall dismissing TikTok's early dancing and lip-sync phase as a fad that would not last.

The participants contrast this with the Chinese version of TikTok, described as curated toward martial arts, scientific discoveries, athletic achievements and children playing the violin, and as locking out users who enter a child's age after 10 p.m. because the government requires it. The host's assessment holds both sides of the tension:

Like, it's not good to censor people, but if you want to have a society that's not captured by horseshit, that's the way to do it.

The guest, born in England to Hungarian parents, offers a parallel from childhood visits to communist Hungary: they recall, while allowing "Maybe I'm getting this wrong," that there was no television on Mondays so people would go out, talk and play sports. A communist regime could impose that at will, and while someone was evidently thinking about the problem, the guest states, "I prefer freedom."

That tension frames a trial described as starting the day of recording, over claims that Meta's platforms addicted children. The figure first mentioned is $1.4 trillion, which one participant doubts will be awarded. The guest says they are torn: they hold strong views on children and technology, but suing large platforms feels wrong because the deeper problem is phones and screens, and another product would take Meta's place.

The guest cites their 2015 book, More Human: Designing a World Where People Come First, which argued that institutions had become too big, bureaucratic and centralized. It also argued that children under 16 should not have smartphones, though they could have other phones to stay in touch. When the participants say YouTube may also be targeted, the host compares that to going after television and notes it would implicate podcasters too. A sponsor read interrupts this part of the conversation.

A lookup read on air clarifies the numbers: reports differ, with recent coverage putting the states' position at roughly $200 billion, while Meta cited possible exposure as high as $1.4 trillion, and the court would decide any penalty. The requested changes are listed as parental verification for teen accounts, changes to recommendation algorithms characterized as dopamine-manipulating, and removing or limiting like counts, infinite scroll, autoplay, appearance-altering filters, multiple accounts and disappearing content such as stories.

The participants push back on parts of that list. Stories, for example, carry worthwhile content such as bowhunter Adam Greentree's daily posts from solo trips of around 28 days in the Colorado Rockies. Their central objection is to the remedy itself:

See, the problem is as soon as the government gets involved in mandating product redesigns, we're fucked.

They agree the problem is real but that bureaucrats micromanaging products is the wrong answer. As an alternative, they point to a reported MySpace relaunch that aims to be algorithm-free, on the theory that engagement algorithms hook users mostly through outrage, whereas a feed of chosen interests would hold only what users want rather than what angers them.

The participants describe these products as very potent, especially for children, and admit to being hooked themselves: "The commodity is your attention." The guest reports that after many campaign events over roughly a year and a half, some with thousands of attendees, voters start with expected issues such as taxes and gas prices. When children and screens come up, though, the mood turns intense and angry. Parents ask what they are supposed to do about screens, social media, video games, and now AI that children talk to and that does their homework.

The host argues video games are far more addictive than social media yet face no comparable campaign, and suggests the reason is that social platforms can shift public opinion. That leads to a contested claim from the conversation, offered without supporting evidence: the rise of the Democratic socialists is attributed to social-media manipulation.

The argument against socialism is put bluntly in the conversation:

It's not natural that people are sitting around, you know, reading Marx and going, what a great idea.

Few online Marxists, it is claimed, understand the historical consequences: once everything is handed to the state, the state has the guns and dictates what people do. The same exchange concedes that capitalism is brutal and needs guardrails. The guest jokes about Mamdani's state-run grocery stores, recalling the only kind of grocery store in communist Hungary, one at the end of their grandmother's block, which they call terrible.

On safety nets, the host lays out both sides: a widow with children and no income should be helped, but such programs can also encourage dependence and "zero ambition." In their words: "I think both those things are true."

The guest says they have long opposed universal basic income, while seeing an argument for a low-level version. The right-leaning versions associated with Milton Friedman and Charles Murray, the guest says, would replace all other benefits and services, including Medicaid, with a simple payment. Both agree that such a scheme might be more efficient because it would sharply shrink the bureaucracy and the number of people controlling others, while noting this is not the version being argued for today.

The host suggests that part of welfare's problem is that it pays too little to live on. The guest connects this to an idea from More Human: in high-cost places such as California the minimum wage cannot cover living costs, even for "seriously hardworking people, often two jobs, who can't live on what they make."

The guest's proposal, the business-friendly living wage, would raise the minimum wage to a calculated living wage that varies by area. They cite figures around 34 in some places when asked about $40 an hour, and say the official low-income threshold in certain California counties exceeds $100,000 a year. The host recalls when $100,000 meant someone was doing very well.

The guest describes a campaign pledge of no state income tax under $100,000. Nationally it drew criticism as too generous, but feedback in the primary said it was too low, since people earning around 120 struggle in expensive areas, so the threshold was raised to 150.

Acknowledging, as a former restaurant operator, that such wages would be hard for businesses, the guest argues that firms already pay payroll and other taxes that the government hands back to low-paid workers as benefits. The proposal is to remove that loop:

why don't you just cut out the middleman and have the businesses pay the workers, not the government?

The guest frames roughly equal net cost for businesses as the goal. The host is skeptical, questioning whether the amounts could really be equivalent and asking what California's minimum wage is. The answer, that it has just risen and depends on location, leaves the question open where this part of the conversation ends.

This stretch of the conversation opens with California's wage floors. The figures given are a general minimum of "16 and a bit" dollars, $20 for fast food workers, and a proposed $25 for health care workers that met pushback. The participants say the $20 fast food wage went through and was followed immediately by job cuts and automation. They give no source for the job-loss claim.

The discussion separates two kinds of work. Entry-level jobs are framed as a training ground for young people, where they learn to show up on time, collect a paycheck and find out how much tax is withheld. One participant puts it this way:

You're not supposed to make a living off of that. You're supposed to be a kid that's learning how to work.