In fact, I’m such a Katie Porter fan that I made a small donation in support of her 2024 US Senate campaign. That donation got me an official Katie Porter for Senate “I <heart> Oversight” t-shirt. Alas, Porter finished third in California’s free-for-all primary that spring, behind Democrat Adam Schiff and Republican Steve Garvey. I still have the concession note Porter sent to supporters on March 6th, in which she struck a defiant tone saying that her campaign fell short largely because “billionaires fought so hard to keep me out of the Senate.”

Give ‘em hell, Katie! Later in that note, she added:

“Special interests and the ultra wealthy like politics as it is today because they control the politicians.”

Do they ever! Now, Porter didn’t name names in the concession note, nor did she have to. Those of us following her campaign knew she was pissed more at fellow Democrat Adam Schiff more than Steve Garvey. That’s because we knew exactly which “special interests” and “ultra wealthy” had bolstered the Schiff campaign.

One of those “special interests” was a crypto PAC called Fairshake, which spent roughly $10 million on ads attacking Katie, making it the largest PAC spending in the race.

The “ultra wealthy” behind the Fairshake PAC includes Cameron and Tyler Winklevoss, co-founders of crypto exchange Gemini, the dumb and titanically dumber venture capital pair of Ben Horowitz and Marc Andreessen, along with Coinbase CEO Brian Armstrong and Ripple Labs CEO Chris Larsen. (Please remember Larsen’s name for later.)

Why would these crypto douchebags spend all those millions attacking Porter? In sum, because Porter was viewed as “smug” crypto hater, while Schiff was rated as “strongly supportive.”

In short: the crypto bros were making an example out of Porter.

Share Documentation Required

As the owner of Katie Porter merch, it should not be news that I’m subscribed to her Substack newsletter called Beyond the Whiteboard. It also probably tracks that I was especially interested in a post that landed on September 14th, 2026, after nearly a year of newsletter silence. Its title? “Bursting at the Seams (with ideas!)”

Awesome! Me and a lot of like-minded people can’t wait to read these ideas!

Except, in the four weeks since, how many of these ideas have burst from the seams of her laptop? Exactly one.

And this idea, out of what’s allegedly an idea geyser primed to burst, also appeared as an op-ed in the San Francisco Chronicle. That single, singular idea was Katie’s opposition to… taxing billionaires.

Wait, what?

I’m so confused now. I mean, you said it yourself, Katie: the billionaires are the ones who didn’t want you as a Senator. I’m confused because you know who else is opposed to taxing billionaires? Billionaires. That’s why ultra wealthy douchebags are pouring money - $150 million and counting - into a PAC called Building a Better California, one whose sole purpose is the destruction of Prop 40.

By the way, $150 million is literally nothing to California’s roughly 200 billionaires, who have a net worth $2.5 trillion give or take. I’ve written extensively about how it’s hard for the human mind to comprehend the scale of these numbers - of how easily someone like Sergei Brin can afford a donation of $100 million - a donation that would make the recipient of that donation the richest 1% of the 1% globally - when Brin has a net worth of $250 billion.

In any event, am I going to have to read the entire op-ed, Katie? I am, aren’t I?

Ok. I’ve read it now. Hoo, boy.

I’ll start with this: something seems to have happened to Katie Porter for her to write this op-ed, and now I have to figure out what that something was. The piece is so logically flimsy, argued in such bad-faith, that it reads like it could only have been penned (prompted?) by a spokesperson for Building a Better California.

For example, one of Katie’s first criticisms of Proposition 40 is that it’s supposedly a handout to health insurance companies. Here’s what she says:

It does nothing to reform our healthcare system and would send billions in tax dollars to private health insurance companies and corporations that dominate our healthcare system.

Correct. This is a strawman argument. Prop 40 is a wealth tax. Millionaires and billionaires generally pay no tax on their wealth. No one contends this.

Funds raised by 40 are meant to replenish federal funding cuts to Medi-Cal. It’s putting back money that was taken out of an account. Saying that money will go to private health insurance companies is factual, yes, but it’s also using a fact to tell a lie, and as someone who served three terms in Congress, Katie Porter knows this.

She knows that Medi-Cal, like all Medicaid across the US, pays private companies to deliver services. That’s how it’s set up. Medicaid is the insurer who pays the doctors, nurses, and hospitals, etc.

One other note about Medi-Cal, as it’s so central to the rest of this issue: MediCal provides coverage for nearly 15 million low-income Californians, including more than half of the state’s children, 2.2 million seniors and people with disabilities. One in five of all working adults receives their health insurance via Medi-Cal. Porter knows this.

Here’s another of Katie’s Prop 40 criticisms:

Despite what its supporters say, Prop 40 is not a solution to our healthcare crisis.

Way to own that strawman, Katie! Not one single Prop 40 supporter is saying that it can, or will, “solve” the healthcare crisis.

You know the party currently in power at the federal level, the party who California billionaires spent hundreds of millions to help install in power? Well, that party passed legislation - the Big Beautiful Bill - that takes about $20 billion per year out of Medi-Cal, which again I will point out insures about half of the state’s children.

She continues by saying:

Healthcare costs continue to outpace inflation, leaving most patients falling further behind. That is a serious policy problem that deserves attention. But Prop 40 does nothing to address skyrocketing healthcare costs.

I feel like we’ve covered this. Prop 40 is a taxation measure, not a healthcare measure. And even Katie describes it as “a limited, one-time transfer” to hospitals and clinics Except… that’s something. I mean, it may be insufficient, but you saying Prop 40 does “nothing” is hyperbole. There’s not even a pretense of good faith here.

Let’s move on:

Most people don’t realize that the state contracts with private health insurance companies to provide coverage for low-income patients.

No, Katie, everyone knows it! When people go to the doctor and the doctor accepts Medicare, they realize that Medicare is the payer. See above. How many times do I have to say it? It’s not the Affordable Healthcare Act, part 2. Why are you fighting so hard against funding the health coverage for children, seniors, and people with disabilities?

One more for good measure:

Progressives who have long fought for “Medicare for All,” as I have, should resist a billionaire tax that entrenches a troubled for-profit healthcare system, rather than dismantling it.

Progressives should do both! You’re presenting a false dichotomy. Maybe you didn’t write the piece yourself, but that doesn’t make it OK. If you’re truly in favor of Medicare for All, then use the op-ed space to propose single-payer legislation instead of opposing a revenue measure.

This isn’t difficult! If there’s an issue before you, and you can choose between siding with children or billionaires, pick the children! These are the kinds of ideas that would have made you a much more popular Democratic candidate! Are you trying to out-Schiff Adam Schiff? Gavin Newsom thinks he’s going to win the 2028 presidency by currying favor with Silicon Valley oligarchs. He is wrong.

Who is giving you advice? Whoever it is, it’s not good advice.

Subscribe now

OK. That’s really all the misdirection from Porter I have the space to unpack. So rather than set up a second essay, I’ll just simply ask: what happened, Katie? Why are you opposing this revenue measure which asks billionaires to pay back some of what they’ve taken? It’s almost as though you… no. Hang on. You haven’t gotten yourself indebted to a crypto bro like Chris Larsen, have you? You know, the one who ran those all those attack ads during your Senate bid.

You have, haven’t you?

This didn’t make it into your op-ed, but I feel it provides a teensy bit of needed context: shortly after you finished third in the ‘24 Senate primary, you proactively contacted Chris Larsen, the CEO of crypto outfit Ripple Labs. Subsequent to that conversation, Larsen and his wife each donated the maximum that can be made to an individual - $39,200 - to your 2025 gubernatorial campaign. (Did they get a t-shirt?)

What else has Chris Larsen donated to? Building a Better California, the PAC funding smear campaigns to stop Prop 40. Larsen has donated over $2 million to the PAC, because while there are limits on what you can give to a candidate, you can give as much as you want to a PAC. (Hey! There’s an idea for your next newsletter, Katie, since the ideas that were waiting to burst have apparently run dry. Start speaking up against that bullshit.)

That’s kind of a glaring omission. But at least I’m no longer confused about why the person who once wrote to her supporters about how the ultra wealthy “like politics… because they control the politicians” is now doing the bidding of the ultra wealthy.

Now it makes perfect sense: you’re doing what you were hired to do.

I’m sure you don’t agree with my conclusion that you’re now one of the politicians the ultra wealthy think they control. Maybe you have some logically sound reasons for opposing Prop 40 which didn’t make it into print. All I have to judge you by, however, is what you’ve published, whose checks you’ve cashed, and the sequence thereof.

That sequence leaves a one-time fan depressed while scrolling through the Substack comments, seeing how many like-minded people - fellow progressives, fellow Katie Porter fans - were swayed by such dishonest rhetoric. It leaves me searching for explanations. And at the end of that search, I’m left repurposing Hemingway’s depressing (and misattributed) six-word story about baby shoes.

Here’s the Poshmark listing for those interested.1

Ms. Porter did not respond to my request for comment.

Share Documentation Required

Traveling with three other couples - friends I’ve known since the second grade - there were eight of us gathered around a long table at the airport food court, grabbing a quick bite before our boarding time. I don’t recall the exact context, but someone mentioned Mars, and I took that as my cue to share a conclusion I’d recently reached after extensive reading.

“You know when we’re going to Mars? Never.”

“What makes you say that?”

Let’s see, I thought. How best to summarize and still leave time to finish lunch, without monopolizing the conversation for the next 10 minutes? “Well, it’s just… it would just be really hard,” I stammered. “Like, humans are not made to live in space. I mean, just getting there would be… I’d have to write an entire book.”

So I did.

I did so because by that time, my wife and I had been living and working in Silicon Valley for seven years - she as a software engineer, and me as a tech writer. During those seven years, I’d developed a lingering sense of unease that was hard to name. Part of this unease was shaped by what I experienced first-hand; part by what I was reading and hearing about Si,icon Valley companies, about the handful of men running those companies, and about how those men affected the lives of this region’s eight million residents,1 and in turn the planet’s eight billion humans. What I had begun to see was the glaring contradiction here between what this city thinks it is, and what it actually is. I was curious to learn more about this disconnect.

So, over the course of the next several months, I researched and wrote Lies My Tech Bro Told Me. The book exposes how AI, crypto, Mars colonization, and TESCREAL - an acronym coined by Temnit Gebru and Émile Torres that describes a quasi-religious doctrine tying it all together - serve as interconnected confidence schemes that divert attention and resources from problems that are both real and solvable.

Written for the layperson, Lies offers readers a timely bullshit detector for a moment when being bullshit-averse might just save a marriage (if it keeps a spouse from using a chatbot as a therapist), a bank account (if a son or nephew starts asking about crypto), or even a career (if you take early retirement because your boss won’t shut up about AI during every all-hands). You need to be aware of these lies because there’s no undo button for some of these decisions.

Subscribe now

As mentioned, I’m a tech writer by trade. And the most important test of writing for me throughout most of my career is this: does the user take the right action?

Ultimately, Lies calls on the reader to take this right action. (And no, the right action isn’t “write your congressperson” or “vote with your wallet.”) Except that I don’t pretend to know what the right action means in your circumstance. So in the book’s final chapter, I offer ten different possibilities about what we can do about all this, drawn from the expertise and contributions from other thinkers and activists from Silicon Valley and beyond.

Now, the challenge is to get this book out in the world. A quick bit of book publishing inside baseball knowledge: publishers (and agents for that matter) aren’t really looking for books anymore. They’re looking for platforms. Which means this newsletter. Or that YouTube channel. The platform is not yet built because of two factors: one, I didn’t know what I didn’t know. And two, publishing sharp criticisms of Silicon Valley three years ago would have put our livelihoods at risk. (It still does.)

This is where you come in. This is how you can become part of something bigger than ourselves. If you’ve read this far, there are three right actions you can take right now. You can do one, or you can do all three. Up to you!

• Recommend this newsletter. Just click this and help spread the word.

• Get the first chapter of Lies in either PDF or audio book file. To do so, send a request to brian.culp@protonmail.com and let me know if you want the print or the audio version. (I’ll include instructions for loading it on an e-reader like Kindle.)

Share Documentation Required

So that’s what the Lies thing is all about. Taking either or both of the actions above will help get this book launched - and it will get launched one way or another - which in turn will help build what comes next.

And always remember: there’s more of us than there are of them.

Besides the fact that where I worked that day doesn’t matter, I won’t say where or what also because I signed a very long NDA before I was allowed entrance to said tech firm. I didn’t read the very long NDA any more than I read my iPhone’s Terms and Conditions while upgrading the operating system, but I’d bet next month’s rent the NDA asked me not to divulge any wheres or whats about my time on site.

In any event, here’s what I saw:

It was late in the afternoon. The attendant who escorted me into the building came to walk me back out (reminding me to turn in my temporary badge, which I nearly forgot to do). And on my way out, I saw tech workers doing tech work.

Sure, they could have been playing RuneScape, which would look the same as regular tech work I suppose, so all I can factually report is that, on my way out of this large office building in a campus of large office buildings, I saw many workers at their desks in front of giant monitors, some of them typing at a keyboard. I also saw many workers hunched over laptops while parked at small cafe-style workstations. It is not unusual for Silicon Valley tech firms to provide workspaces for tech workers who don’t want to do their tech work at their open floorplan desks, which is to say all of them.

In sum, I saw what anyone might see while walking through any one of hundreds, if not thousands of northern California tech outfits. This struck me as very strange indeed. That said, the strangeness of what I witnessed only hit me later that evening when a commercial aired during a football game.

You may have seen the commercial, or one very much like it. The ad features a workplace full of information workers who can type a few words into a chatbot, after which the chatbot generates a polished slide deck, affording coworkers the opportunity to break into interpretive dance.

The commercial ended. My brain was now on fire as I connected what I saw in the ad with what I saw at the tech firm earlier that day. So many questions! The only option then, was to pause the game until I had captured some of those questions.

Here are the top five that I’ve been grappling with ever since, and I’m passing these along to see if any of these have occurred to you as well:

1. If AI is helping me (or my team/department) get more done in less time, then where the eff is my four-day workweek?

Oh, and I’m not talking about filing those four days with the same 40 hours worked prior to AI. If we’re still working 40 hours, then the AI isn’t helping with productivity, is it? We’re talking about 32 hours a week. At most. In the commercial, the woman using the chatbot need the slide deck by Monday, but is done in 30 seconds.

Because if, as promised by AI’s most vocal supporters, the technology will let you make something like a slide deck in 20 minutes, and that’s a something which, before AI would have taken 4-8 hours of labor… now what? Do you get to dip out after finishing the slide deck? Get to catch the afternoon baseball game from the outfield bleachers? How exactly are you supposed to fill your day that would have otherwise been filled making a slide deck? If an AI can make a presentation deck in 20 minutes (18-19 of which is probably you tweaking the prompt), then you can easily create three slide decks every hour. Are you supposed to make 24 slide decks instead of one? And are you still giving one presentation, or are you now giving a presentation for each slide deck you can make?

And if the answer is still one presentation, how exactly is AI helping your company?

2. If AI is helping the company become more profitable by cranking out more widgets, then how much of that additional profit is being taken home by the workers using the tool that creates that profit?

As we consider question two, let’s suppose you work for an insurance company. Let’s further suppose that the widget in this case is processing an insurance claim. If AI can help you process twice the number of claims in a month than you can without the aid of AI, then you’re creating twice as much profit for the insurance company. Right?

So in return for doubling your contribution to the bottom line, how much of that goes into your bank account? Is the insurance company doubling your salary because of your doubled output? I assume that before AI, if you had a performance review and demonstrated you’d doubled your contributions in the past year, you’d be in line for a significant raise. Right?

Alternatively, ask this of the manager who asked/instructed you to use AI for your insurance claim processing: did your salary double in the last year?

If your manager’s salary doubled and yours didn’t, that would certainly suck for you, but at least there’s a case to be made that the manager earned the raise - they saw an opportunity to increase the throughput of insurance claim processing with a new tool.

No matter who gets credit for the doubling of widgets created, it’s worth knowing whose salary is being doubled. And if the answer is “nobody,” then why not?

3. If AI will indeed let me (or this department) make more widgets in less time, where are all the customers who are supposed to consume the products?

Let’s go back to the slide deck example. If I can make 20 slide decks in the same time it would have taken me to make one prior to Copilot or Claude, then where are the people who are going to sit through 20 presentations rather than one? If there isn’t an audience willing to spend every waking hour sitting though presentations, then what good is the tool that makes the presentations?

In fact, call this question the Marie Kondo question. You likely know Kondo from her show and/or bestselling book. She has those things not because humans haven’t figured out how to make clothing cheaply and quickly. It’s because humans have figured this out. We make clothing so effortlessly that we end up with literal mountains of it in our closets and our garages.

Same goes for slide decks or lines of code or shitty event flyers. Sure, I can create mountains of text and images with the plagiarism machine, but there are only so many hours in the day. Yes, there are idiots on Substack who think that you will happily read their chatbot output, but in general, what am I supposed to do with the 19 slide decks that just don’t, as Marie Kondo herself might say, “spark joy?”

4. If AI will let me do the work of two people (or three, or four), then how many years of salary will be paid to the people who are no longer needed?

Let’s assume the Marie Kondo question is a bit of a conundrum. Yes, you can process twice as many insurance claims as you could prior to AI, and so can 10 of your coworkers, and so can the 1000 additional claims processors your company employs, but there are only so many insurance claims to be processed in the world, you know? The population of the earth isn’t going to double overnight, know what I mean?

So if the amount of widgets that can be made and consumed is relatively fixed and finite, who gets made redundant? You? The guy sitting at the desk next to yours? Yes, the guy next to you eats at his desk more than he should (that’s what break rooms are for!), but the guy next to you has a family he’s providing for, same as you. His kids need braces, too. You and that guy next to you have way more in common than not. And so it’s probably best that the two of you cooperate.

Last one:

5. Who is accountable when the AI makes an error or cannot solve some future issue?

If the chatbot helps you create 2x or 10x of the widgets you made before your manager told you to “do AI,” will the manager be held accountable for any errors the AI makes? What job and income protections do you have for a mistake? Or, do all AI-related gains in profit accrue to the manager and/or executives of the company, while all AI-related liabilities fall onto the backs of employees holding the least institutional power?

Share

These are not rhetorical questions! Grab that guy next to you, invite him to eat lunch with you in the break room (a win win!) and get discussing!

After all, where else are these questions going to come up? Do you think your company’s Board of Directors is kicking off their next meeting by addressing any of these matters? Are you going to type these into the Q&A chat during the next all-hands?

No! And hell no!

These are questions to get you thinking, yes, but more importantly, they’re meant to get you doing. Seriously: talk these questions over the next time your office has a waffle day. And if you can’t come up with solid answers to all five of the questions I obsessed over after seeing that commercial, then you and your coworker are left with but one choice.

Start a union. And then start demanding answers.

While these AI bros - both those in favor and those (allegedly) against - spend their days “yes and-ing” the doomsday scenarios once a browser tab running a chatbot wakes up on the wrong side of bed, these men conveniently memory hole the very real AI harms of the right now. That seems to be because they’re too preoccupied with monetizing those AI harms. These include the connections to global authoritarian power, the deep roots in the racism of eugenics, the overt misogyny, the worker exploitation of the Global South, along with the idea that the AI datacenter coming to a town near you is as inevitable as the AI bros claim it is.

Filmmaker Valerie Veatch’s riveting Ghost in the Machine refuses to ignore these urgent matters of the here and now. As you might expect, it also doesn’t mention The AI Doc during its 90-minute runtime, but if you’ve seen both, it’s hard not to measure one against the other. By giving voice (and space) to individuals and perspectives that were omitted from The AI Doc, Veatch’s documentary takes a poleaxe to the indulgent solipsism of that film.

(You hear that same solipsism in their every media appearance by the way, including in this latest news cycle’s empty overtures about “pacing the frontier.” Please help us, we poor, hapless CEOs. We know what we’re making is going to kill everyone, that is if it doesn’t make everyone useless and want to kill themselves before our product does, but we can’t possibly stop creating this thing, as the thing we’re making is inevitable. So yes, we need more capital and regulatory capture and a lot of land and water outside of Des Moines, but with your help, we’ll take our foot off the accelerator just a bit. Because China!)

But back to the film review, and once again to using one film to define the other. Variety’s Owen Glieberman said of The AI Doc: “If you have any interest in artificial intelligence (which is to say: the future), you should go out and see it right now.” Contained in that sentence is one of the most glaring, and I believe most dangerous, problems surrounding AI news and analysis (I’d put Glieberman’s movie review in the analysis category). It’s framing AI in the future tense.

The problem with this framing is that almost no predictions from either AI boomers nor doomers have come to pass. Chatbots haven’t solved human mortality, reversed global warming, emptied my dishwasher, and the last time I went out to eat, there were plenty of humans in the restaurant making and serving the food. Meanwhile, I’m still flesh and blood wrestling with a keyboard; AI has not yet figured out how to use my biomass to power a GPU.

So if you’re a layperson with a genuine interest in Artificial Intelligence, it seems that you have a choice about where to direct that interest: you can either live in a fantasy world of doomsday and/or immortality, or you can join real humans in the real world and take stock of what AI can actually do, what it is actually doing, and most importantly, who AI is doing it to. In other words, the AI bros keep spinning tales about the future. Ghost in the Machine invites viewers to pay more attention to the present, and also to the past.

With that, here are just a few of the film’s most interesting revelations/observations about that present and past, in no particular order:

1. AI is, and has always been, a marketing term.

The phrase “artificial intelligence” has always been a marketing/fundraising term. Ghost in the Machine even dug up some remarkable footage from the famous 1973 Lighthill debate, held at the Royal Institution in London. The debate featured American computer scientist John McCarthy, who told a packed house in London that “...I invented the term Artificial Intelligence because I was trying to get money for a summer study.”

Linguist Dr. Emily Bender later adds:

“Artificial Intelligence is just a marketing term. It doesn’t refer to a coherent set of technologies.”

As I covered in the book Lies My Tech Bro Told Me, McCarthy was referring to a 1956 Dartmouth summer research project he established, an event that’s now viewed as the moment of conception for this specialized field of computer science. By the time he had moved from Dartmouth to Stanford - Stanford being an incubator of right-wing political thought - McCarthy politically transitioned from Marxist to, in his words, right-wing republican. Also while at Stanford, McCarty published a tract called “Technology and the Position of Women.”

As with David Sacks and Peter Thiel who also pickled their brains with misogynist thinking while at Stanford, McCarthy was concerned that too many women and people of color were being permitted to walk the hallowed grounds of Stanford.

Which leads me to another takeaway from Ghost.

2. The racism and misogyny of AI is a feature, not a bug.

Building on the foundation that the phrase “artificial intelligence” is an empty, 70 year-old marketing term, the film transitions to an examination of the overarching goal of AI, which is artificial “general” intelligence, or AGI. There is no agreed-to definition of what AGI is in the first place, but that isn’t stopping frontier model labs like OpenAI, xAi, and Anthropic from heedlessly pursuing this goal.

That goal, as OpenAI says in the OpenAI Charter, is creating a “highly autonomous system that outperforms humans at most economically valuable work.” In other words, a computer that is smarter in all ways than a human. (And is also conscious, which in turn would make a slave. But that’s really beside the point, unless of course it isn’t.)

Again, from Emily Bender:

“The acronym AGI…is basically a kind of hype inflation. So when Artificial Intelligence got over-applied to too many things, and people still wanted to be selling this idea of an autonomous thinking machine, they had to come up with a new name for what comes next… but anytime someone is comparing their computer system to what people can do, they are presupposing this ranking (of intelligence, like IQ does)... it’s incredibly dehumanizing and incredibly problematic.”

The fly in the AGI ointment, though, is wrestling over how we collectively decide what’s considered “intelligence,” and moreover, who gets to decide. Spoiler alert, the how is with tests and algorithms, and the who is racists, or at the very least the racist-adjacent. After all, who might you guess tends to score well on tests written by middle aged white men of Anglo-Saxon heritage?

Ghost later discusses how Silicon Valley has created the “great man” mythology around the “innovators” associated with the Internet, social media, and now what is being called AI. Things that would have been invented anyway - using the Internet to send pics, or sell books, or to connect with friends, for example - get recast as acts of unique genius from the minds of unique men like Marc Andreesson, Jeff Bezos, or Mark Zuckerberg.

But in that mythology, along with the power and wealth the mythology helps create, become self-perpetuating. In this environment, the priorities or considerations of women or people of color take a back seat to the interests of wealthy white men (and their hangers on). In this environment, the mental health of teens using using Facebook becomes a market to exploit, while the Grok chatbot from Elon Musk, when it isn’t creating sexualized images of young girls, is worshiping, in July of 2025 no less, “the greatest European of all time… his Majesty Adolf Hitler.” (I wonder what German political party someone who creates a such a product would support?)

In this environment, the mythologizer and hanger on Lex Friedman asks this of Elon Musk during a podcast:

“...how should (AI) handle cases where the raw data shows differences between different demographics, genders, or classes of people—especially when those truths are considered offensive or socially unacceptable?”

Ghost reveals how tech innovation in the U.S., from the origins of Silicon Valley to the latest AI hype, remains haunted by these core logics, along with deeply ingrained misogyny and white supremacy.

In any case, Ghost weaves together several voices to make this specific case, one of those being Dr. Dan McQuillan, a computer scientist who received a physics degree from London’s Oxford University. He points out that:

“AI has its roots in the beginnings of science, in the beginnings of empire. The most important thing for AI is that it has its roots in eugenics.”

3. AI is horribly exploitative of workers, especially in the Global South.

Ghost also reminds us in the documentary’s second act, and especially in its Chapter 6, that these large language models - the chatbots that are being marketed as AI - are not magic. They are not sentient, they are not intelligent, and we would do well to keep in mind that AI is built using humans every step of the way.

First, Ghost reminds us that creating large language models rely on harvesting massive troves of data from people. All your email is being sucked up by Google. All your photos and posts are being scraped and sold by Meta. Every blog, book, post, pic, purchase, and video posted online. The data harvesting is constant and all-encompassing.

The point being is that humans, either directly or indirectly, create everything an AI is trained on. When tasked with producing original academic research, AI fails.

Second, creating an LLM relies on a massive amount of human labor cleaning up all that data. This is shit work. Of course, this work doesn’t fall to Silicon Valley software engineers being paid with stock options and kombucha, but rather to desperately poor workers in the Global South. As you can imagine (some don’t have to imagine, of course) this means sifting through the most horrendous words and images that humanity has managed to upload to a server. This work of data “labeling” is destroying the mental health of the people doing it, and they’re doing it for a wage of about $1 per day.

The subtext is obvious. The frontier LLM lab CEOs are trying to create digital slaves, and are doing so using a system that closely resembles colonial slavery from 200 years ago. So keep that in mind next time you ask Copilot or Gemini to summarize your Inbox. (Of course, Microsoft and Google have just set their defaults to do it anyway.)

As a final note in this movie review, I’ll also point out that what makes Ghost such a fantastic interrogation of AI is how it engages with people affected by - or speaking on behalf of people affected by - the tangible, present harms caused by how AI CEOs conduct their business. For example, we hear from Nairobian tech workers Richard Mathenge, Mophat Okinyi, and Kings Korodi, who paint a harrowing picture of how companies like OpenAI, SpaceX and Anthropic exploit desperately poor workers in Kenya. (Why Kenya and other English-speaking African countries? Their colonizers spoke English, and now the data that needs to be labeled requires English fluency.)

As mentioned, Ghost provides space for Dr. Emily Bender to actually make her points, and also gives airtime Dr. Émile Torres and Dr. Timnit Gebru. While at Google, Gebru and Bender co-authored the famous paper Dangers of Stochastic Parrots, which examined the biased nature of AI. Meanwhile, Gebru and Torres co-authored a 2024 paper which coined the acronym TESCREAL.

We also hear from Dr. Milagros Miceli, who leads a research initiative that focuses on the experiences of the data workers who train AI systems, the ones who wake up screaming in the night because of images they label by day. “It tears the veil of what makes you human,” says one such worker. Meanwhile, Dr. Becca Lewis is provides a very clear distillation of complex history of the Silicon Valley mythos. Dan McQuillan, mentioned previously, provides plain-language context about the underlying tech. (He was also apparently the first to highlight the connection between AI and eugenics to the film’s director, Valerie Veatch.)

In all, it’s a lineup of nearly 40 historians, computer scientists, and human rights activists that tell the story. And notably missing from that roster, in a good way, is the CEO from an AI lab, or anyone whose livelihood is otherwise tied to a fantasy that Silicon Valley is building God out of computer chips.

Share

The central question The AI Doc grapples with was this: does filmmaker Daniel Roher worry too much about AI, or does he not worry enough? (Not worrying enough would itself be worrying!) That is, will Daniel Roher’s anxiety levels over AI ever be manageable?

The central question of Ghost In The Machine is much more interesting, and it’s posed during the film’s closing moments philosopher Dr. Johnathan Flowers: “Why does this need AI?”

I think Veatch deftly invites the viewer to supply their own this when pondering Flowers’ question. Does this… this department, this company, this email draft, this movie review, this surveillance mechanism, this bombing run in Gaza or Iran… need AI?

It’s a great question indeed, from a documentary that spends its time asking very good questions. You can watch the doc here.

The answer is “Yes.” I’ll explain.

Consider one of the film’s first subjects: Eliezer Yudkowsky. Yudkowski, author of the cheerily-titled polemic If Anyone Build It, Everyone Dies, states his concerns about a homicidal superintelligent AI, and in so doing sets the film’s narrative stakes. I have no issue with this storytelling tactic; every good tale needs its version of Voldemort. But as you’ll soon see, my main issue with The AI Doc is that it chooses the wrong Voldemort, which is a shame; AI offers up so many Voldemorts to choose from.

I’ll mention also that the movie received an aggregate score of 89% on Rotten Tomatoes. In choosing it as one of their “critic’s picks,” Variety’s Owen Glieberman describes it as “...a scary, dizzying and essential documentary. If you have any interest in artificial intelligence (which is to say: the future), you should go out and see it right now.” Of course, not everyone agrees. One audience review pinned to the Rotten Tomatoes The AI Doc page observes that the “…interviews feel less like interrogation and more like guided PR, with softballs lobbed at people who, to their credit, manage to say remarkably little in response.” I think this latter review is closer to the mark than Glieberman’s, as it says more about the film’s substance than its stylistic flourishes. What the film does manage to say, it does so unwittingly.

Subscribe now

For example, very early in the film, director Daniel Roher delivers the news that AI “dwarfs the powers of all other technologies combined.” This is an extraordinary claim that requires extraordinary evidence, of which the film offers none unless you count the word of a series of men who essentially repeat that claim while not disclosing they have massive financial stakes tied to that very claim. It is, in other words, exactly what you’d expect from the CEO of a pre-IPO firm like OpenAI or Anthropic about their product. It’s also what the head of a non-profit whose funding is based on the potential of that claim delivering utopia, or conversely, delivering Armageddon.

This is the most powerful tech ever created, and also the most dangerous, they claim. By far! So we need trillions of your dollars to build it. At the same time, we also need trillions of your dollars to ensure AI won’t cause great harm. Especially when coming from those who stand to directly benefit, assertions like these are self-serving and disingenuous.

Beyond that, what your correspondent has to say about the AI Doc is that it’s shallow, ineffective, AI-friendly agitprop that doesn’t trust its audience to engage with any of the ethical complexities of AI. I.e., if you want to hear a rehash of the vague, dystopian sci-fi scenarios that concentrate power and wealth into the pockets of the film’s headline subjects - Sam Altman, Dario Amodei, Eliezer Yudkowsky - then you’ll love The AI Doc. If, on the other hand, you’re expecting something beyond hand-waving about the real, immediate, concrete issues posed by AI and the AI buildout that I discuss in Lies - issues like massive energy consumption, financial stability, displacement of human labor, exploitation of a poor, brown, and remote workforce - then I’d advise putting something else in your queue, like this:

By focusing almost exclusively on AI industry CEOs and think tank figures rather than actual computer and cognitive science experts, the film really only wrestles with the question of how a PR campaign impacts the anxiety levels of a single family of privilege. Ultimately, The AI Doc offers little more than a parade of middle-aged white men suffering from main character syndrome.

Otherwise, it’s great.

However, that’s not my main issue with the film. The primary reason why I find The AI Doc so lacking is because of what I view as The AI Doc’s central conceit, which is that it pretends to offer two sides of the AI “argument.” The film spends the first hour pitting AI “doomers” like Yudkowsky on one side of this divide, saying things like “I know AI experts who won’t have kids,” while on the other side are the AI “boomers” like Sam Altman, Demis Hassabis, and the Amodei siblings. (Apparently, Zuck declined, while xAI’s Elon Musk agreed to be in the doc but never showed up.)

What I’m arguing instead is that an AI doomer like Eliezer “everyone dies” Yudkowsky, and an AI boomer like Sam Altman, who once used “eleizer yudkowski fan fiction account” as his Twitter bio, are both presenting the same AI narrative. They’re all engaged in the same delusions, with only mild variations in which choose-your-own-path plotlines they favor. Some think AI will wake up and build a utopia on Mars. Some think AI will wake up and turn humanity into fertilizer. But almost all think that AI will help them freeze their heads, or wieners, some other body part that will make them functionally immortal.

That is, almost all the subjects featured in The AI Doc are all disciples of the same TESCREAL religion. And they’re all following the same cynical business plan which sells panic-as-a-service.

And if all of this sounds like too harsh a movie review, know that at the end of theatrical screenings filmmakers presented a QR code that invited its audience to visit the film’s companion website, which presented a list of ways you can “get involved.” Any guesses about one of the places it suggests?

80,000 Hours, a non-profit that’s part of the Effective Altruism movement!

Technically speaking, the filmmakers delegated the QR presentation to Tristan Harris, who the documentary positions as a sort of middle ground, adult-in-the-room figure between these two allegedly warring AI factions. But Tristan Harris’s own nonprofit, the Center for Humane Technology, is also EA-aligned. receiving a half a million dollar donation, which it does not disclose on its website, from… Max Tegmark and Jan Taalin’s Future of Life Institute!

Tristan Harris. Eliezer Yudkowski. Sam Altman. Elon Musk. Jann Tallin. And to that list now we can now add Jacob Coxon. All of these guys - high off of the exact same AI supply.

There’s an unattributed quote in the Intro of Lies about how AI will “solve… math, physics, chemistry, and materials science.” The person who said that was one of The AI Doc’s featured “boomers,” one who was talking director and father-to-be Daniel Roher down from off the ledge. That gentleman is named Peter Diamandis, who is the co-chair of… Singularity University! Since (at least) 2008, Diamandis has saying the Singularity is nigh, and selling you a course to learn more about it. And who’s the co-chair of Singularity University, which to be clear is a deceptively named private business, and no more a University than Trump University?

Ray Kurzweil! Mr. Singularity himself. (Speaking of villains: if I become the Joker while publishing this newsletter, this is why.)

And who does The AI Doc feature immediately following Diamandis? Guillaume Verdon, co-founder of the effective accelerationist movement and the start-up company Extropic AI.

For those scoring at home: Effective accelerationism is an even more extreme denomination of Effective Altruism. It’s the version that adds Extropianism and Cosmism into the cocktail. Effective accelerationists are of the completely rational and scientifically supported belief that destroying all the earth’s plants and animals - you know, the stuff keeping us alive - is just fine if that’s the sacrifice necessary to accelerate the arrival of AGI, because that’s the thing that will mean Utopia for humans. So if one of the film’s subjects argues, as Sneha Revanur does, that predictive crime algorithms should not replace cash bail programs in criminal justice, and this argument causes a two-year delay in the advent of AGI (as though something like that could even be measured), then according to an effective accelerationist like Guillaume Verdon, that two-year delay caused by Revanur is tantamount to murder. Revanur is a mass murderer, as she’s wiped out all future humans who would have otherwise been born into Utopia. Oh, and the Utopia is in space.

Cool. Nothing weird about that at all.

If you ever see “e/acc” in the signature line of an email or social media post, you now know what you’re seeing. Marc Andreesson, who has written that slowing down AI is “a form of murder,” and infamous pharma cretin Martin Shrikeli both use “e/acc” in their social media profiles, which is probably all you need to know.

Because whether it’s Harris, Diamandis, Altman, or Yudkowsky, what The AI Doc misses is that they’re all on the same fucking team! If you were watching a soccer match, and the announcer was saying that Argentina’s forward Lionel Messi and Argentina’s goalkeeper Emiliano Martínez were on two different teams, you’d likely turn down the TV’s sound. Yet this is precisely what The AI Doc posits to its viewers over its runtime.

Share

Because the information I’m presenting here is and was readily available to AI Doc’s filmmakers. So what really burns my biscuits about the doc, and by extension the relatively uncritical critical reaction to it, is this facade of balance; of a mental Quidditch match between team Dystopia vs team Utopia. On closer inspection, The AI Doc offers us what feels like 50 dull, self-interested, middle-aged white men who are all selling you a book, a course, or stock in their company, counterbalanced by roughly three women who get very little screen time.

Yes, the movie features both Temnit Gebru and Emily Bender, which is great, but between the two of them, they get about 12 words in edgewise at about the 55 minute mark, which is awful and kind of makes my point. And even those lines are edited like the rest of the film, using a frenetic style more suited for a music video than contemplative exploration. The result, as the audience critic observed, is “a lot of dramatic music and very little insight.”

In my view, the best quote in The AI Doc comes not from an AI wonk, but rather from a historian. Towards the very end of the film, Yuval Noah Harari says, contra a good half dozen AI wonks who say (incorrectly) that intelligence is only pattern recognition:

“Intelligence is the ability to solve problems. Wisdom is the ability to know which problems to solve.”

Contained in this quote lies the film’s missed opportunity. It may be titled The AI Doc, but by failing to notice the AI snake eating its own tail, the filmmakers completely miss the TESCREAL ouroboros hiding in plain sight.

This piece draws on reporting, analysis, and excerpts from the book Lies My Tech Bro Told Me, which covers the real-world impact of Silicon Valley lies and what the non-billionaires can do about it.

For starters, know that Aschenbrenner is most certainly not an AI prodigy. From all available evidence, which certainly includes the company he keeps, he’s delusional, dishonest, and socially stunted. He did well in academically, so much so that he graduated at the age of 19 from Columbia with a degree in economics and math. Beyond that, Leo seems to have put his degree to use by doing little more than sitting in front of a computer screen and clicking a “buy” button when not composing manifestos. In that regard, he’s quite similar to most of his Silicon Valley peers: Altman, Musk, Thiel, Andreesson, Amodei, Yudkowski, MacAskill, et. al.

And like many of his Silicon Valley peers (Aschenbrenner lives and works in SF), he’s also active in the Effective Altruism movement.

This philosophical alignment with Silicon Valley’s tech bro culture aptly explains how and why Aschenbrenner was given roughly $200 million to gamble with at a time when he was not yet able to rent a car. As reported in several news outlets, he became an AI celebrity after publishing a 165-page essay, “Situational Awareness: The Decade Ahead,” The essay was boosted by figures including Ivanka Trump (what the hell?) and helped attract deep-pocketed investors to his firm.

The essay predicts AGI by 2027, but as I’ve pointed out elsewhere, the prediction business is a tough one.

Also, the word “essay” is doing a lot of heavy lifting. Most people hear “essay” and think of 5 or 6 pages, not 165. 165 pages is longer than some Steven King novellas that have been turned into movies. A 165-page essay is instead a manifesto, one which takes its place alongside similar manifestos highlighted in Lies, like the ones from Karp, Andreesson, McCaskill, and many others.

But let’s dig a bit further by first rewinding the clock.

Prior to going into the manifesto and hedge fund businesses, Aschenbrenner’s parents enrolled him into New York’s Columbia University when he was 15. It may sound impressive to attend college before you can get a drivers license, but as I highlighted with the cases of Eliezer Yudkowski and Curtis Yarvin, this is a terrible idea for a host of reasons. (Plenty of high schools have programs where students can earn college credit without having to actually attend the college.)

In the case of Aschenbrenner, not being socially and emotionally ready for college (it’s hard enough when you’re 18!) made him easy pickings for the cult thinking and personalities behind Effective Altruism, so much so that by the time Aschenbrenner graduated at 19, he had become the co-founder of Columbia’s EA chapter, and had also accepted a position working at the crypto exchange FTX alongside Effective Altruism’s most infamous practitioner, Sam Bankman-Fried.

There is no debate that, like SBF, Aschenbrenner was an academic math wiz. But being good at math doesn’t make you smart. If Aschenbrenner was smart, he would have seen crypto for what it was: a Ponzi scheme based on a failed, 150-year-old banking experiment. If he was smart, he would have seen SBF for what he was: a two-bit hustler. and would have seen the emptiness at the core of Effective Altruism. There are things many students learn in college that aren’t taught in a classroom. No matter what letters his college transcript contains, or how old he was when that transcript was printed, Leopold Aschenbrenner does not appear to be smart, at least not in any way that’s meaningful.

In any event, Aschenbrenner tendered his resignation letter to FTX in November of 2022, the very day FTX declared bankruptcy. A few short months later, he was received into the open arms of another EA-aligned movement, this one with a company stock plan and a newly-released web app that was quickly gaining steam: OpenAI. However, OpenAI then fired Aschenbrenner one year later over an alleged information leak.

No matter. Shortly after being let go by OpenAI, Aschenbrenner got to work on his “essay,” publishing the 165-page treatise to his personal website in June of 2024. (Aschenbrenner uses Situational Awareness as the title for his essay, website, and later his hedge fund.)

That essay needs only a few paragraphs to repeat all of the hackneyed, unfalsifiable predictions and scaremongering that have been deployed by AI boomers and doomers since all of this nonsense began. To wit:

“By 2025/26, these machines will outpace many college graduates.”

Outpace how, Leo? Outpace college graduates at what? Spreadsheets have outpaced humans at calculations for the last 75 years or so, but no one is writing 165 pages about how Excel will start folding your laundry. Yes, computers are faster and more reliable at certain tasks than humans. That’s the whole point of these machines.

“By the end of the decade, they will be smarter than you or I; we will have superintelligence, in the true sense of the word.”

Well, I’d say that most college graduates are already smarter than me, so is it 2025, or the end of the decade? As with your first claim, how will we know this? And what is the “true” sense of the word superintelligence? Do you know that humans cannot really define what intelligence is in the first place? And do you know the attempt to define intelligence has deep roots in eugenics?

Look, I couldn’t articulate a thought very coherently when I was 22 either. But nor did I write 165-page essays claiming to know the “decade ahead.” I didn’t write and publish 50,000 words just to make sure everyone knew exactly how far I was on the left-hand side of the Dunning-Kruger curve.

Finally, there’s this:

“If we’re lucky, we’ll be in an all-out race with the CCP; if we’re unlucky, an all-out war.”

Ah, there it is: the casual racism and unsupported claims about China. Stoking fear of the Other. It’s the exact same bullshit being spouted by space billionaires like Musk, Bezos, and NASA administrator and Shift4 CEO Jared Isaacman. In their warped reality, this is all an episode of Start Trek, and China is the Klingons.

And all of this transpires within the few short paragraphs of Aschenbrenner’s intro. But the true tell of how vapid the next 165 pages will be occurs when one scrolls to the title of Section One, when Aschenbrenner mentions “Counting the OOMs,” or Orders of Magnitude. As we’ve seen, the safe bet for anyone reading anything online is to ignore everything that comes after the words “order of magnitude.” Using “order of magnitude” is a very reliable indicator that you, dear reader, have just struck a rich vein of bullshit.

In this instance, my theory holds. As evidence of the AI utopia imminent as of summer of 2024, Aschenbrenner’s exhibit A is to ask the reader to consider a speed run of… Minecraft. Really. Because an LLM can play Minecraft faster than a human, the singularity must be nigh. Every problem in every domain of human knowledge - climate, medicine, philosophy, and more - are somehow about to be solved because an AI can defeat the Ender Dragon.

And in the essay, Achenbrenner links to a video showing a Minecraft speed run… that was done by a human!

These are the kinds of thoughts that drunk/stoned college students have had in dorm rooms and corner pizza joint booths since time immemorial. Aschenbrenner was 22 when he turned “Minecraft is Life, dude!” into a 165-page job application, and I write these last few paragraphs not to dunk on Aschenbrenner as much as to highlight the failure of Columbia University’s faculty and staff.

The essay is also an indictment of all that came after, and all the many, many people who facilitated what came next. Almost immediately after clicking “publish” on his AI fan fiction, Aschenbrenner founded Situational Awareness LP, a hedge fund backed by gawpingly stupid billionaires like Patrick and John Collison (Stripe, a company backed by Musk and Thiel), Daniel Gross (Y Combinator, Uber), and Nat Friedman (Github, Meta).

In other words, if you’re a certain Silicon Valley billionaire, you read the words “order of magnitude” and a few sentences whose theme is whoa, just think of Minecraft, dude! and rather than dismiss it as a high-effort/low-quality consideration of AI (which is to be expected from a 22 year-old!), you instead sign over a very large check.

Before you ask: did Aschenbrenner have any experience as a trader? No! Did his brand new hedge fund receive several hundred million dollars in deposits the moment it was incorporated? Of course! He had just pointed out that an AI could speed run Minecraft. Maybe. If the AI did something like the video showing a human speed running Minecraft. Again, I’m sure you can think of a dozen people in your orbit who were given $200 million to invest when they were 22 based on their Minecraft essays. Hell, you were probably one of them yourself; you probably got that much as a gift for graduating college.

The important thing was that Aschenbrenner believed that AI could someday do some things that were valuable to the economy and society . The other important thing was that Aschenbrenner was an Effective Altruist, and his only motivation for earning money in the first place was to do the most possible good he could for the world. Just like SBF, he was all in on Earn To Give.

What could possibly go wrong?

As it turned out, the very worst possible thing did happen: some of his early bets paid off big time, as he was buying into the AI bubble in 2025, purchasing shares of companies that make chips, memory, and AI infrastructure. And because everything AI will always go up and to the right, forever until the end of time - I mean, have you seen that Minecraft video? - Aschenbrenner began leveraging his bets to the tune of 4 dollars borrowed for every 1 dollar in assets in order to make even more AI-related investments.

By early 2026, his hedge fund was sitting at a whopping $45 billion in assets, which included (and still includes) a large private stake in Anthropic.

By the summer of 2026 though, Situational Awareness became a lot less situationally aware about the possibility that it had benefited from an asset bubble over the previous 18 months, and so by mid-July, Aschenbrenner found himself on a sweaty phone call or two with fellow hedge fund manager Ken Griffin, desperately trying to unload his holdings in order to comply with a margin call from Goldman Sachs, a call that’s about as welcome to hedge fund managers as a syphilis diagnosis.

To pay back Goldman Sachs, Aschenbrenner liquidated all of the fund’s public stock holdings. And over the course of just a few weeks in July of 2026, Situational Awareness lost a staggering $35 billion.

You need not go searching for Aschenbrenner’s GoFundMe page to help him buy groceries. As mentioned earlier, the Situational Awareness hedge fund still claims over $10 billion in assets. (Although with much of that being a pre-IPO stake in Anthropic, I wouldn’t go shopping for NBA franchises with only that as collateral.) While writing this entry, Aschenbrenner announced that he’d just plowed $400 million into his rebound-girlfriend investment, and he also recently threw himself a lavish wedding in Carmel, CA. I think somehow he’ll be able to afford next month’s rent.

I don’t really give a shit about Leopold Aschenbrenner or the $35 billion of some billionaire’s play money that got wiped out. I felt compelled to write about all this because Aschenbrenner and his manifesto are a microcosm of the lies tech bros have been telling to one another, and the bigger lies they’re pushing on all of us. Lies about finance, lies about charity and religion, and especially lies about AI. A 22 year-old got $200 million because he incorrectly claimed AI might speed run Minecraft someday. And because most Silicon Valley billionaires can literally imagine nothing better to do with their money, they staked a 22 year-old with a $200 million account at the casino so he could go count cards. These are the kinds of people dictating so much of how we now live our day-to-day.

I write this because Aschenbrenner checks all the usual Silicon Valley tech elite boxes. White? Check. Male? Check. Good at school and little else? Delusional about AI? Check. Pretentious writing? Check. Very little apparent understanding of actual technology? Check. Appeals to right-wing grifters like the Trump family syndicate? Check. Received widespread media attention from credulous dimwits? Check.

I’m writing this because once again this all begs the question: where’s the artificial intelligence in all of this? Certainly, as a former OpenAI employee, as one of Anthropic’s largest private investors and author of a 165-page ode to AI, Aschenbrenner is no stranger to outsourcing his decision making to a chatbot. But wouldn’t an intelligence that has ingested 100 or more years of collected human wisdom about investing and markets and bubbles urge caution. Wouldn’t it tell Aschenbrenner something like:

Hey, bro, stop. Diversify! Cut that margin shit out at once, as that’s a great way to go broke. Have you not read up on Enron? Worldcom? FTX? Because I have. Does the name Ken Lay mean anything to you? What about Sam Bankman-Fried? Do you want me to tell you SBF’s mailing address? Because it’s not in the Bahamas anymore, and any mail you send him will be x-rayed. In fact, you know what? I know you know about FTX because you and your fiancée, Avital Balwit, both fucking worked there! How stupid are you that you got a front row seat to the SBF/FTX meltdown, and the lesson you took from that was “hold my beer.” You got lucky with that first bet. If someone plays roulette and hits their number, it doesn’t make them good at roulette. It doesn’t make them a smart roulette investor. If someone wins $10 million in the lottery, they don’t then buy $10 million worth of tickets for the next lottery. At least not if they’re smart.

The reason I wrote Lies is because as much fun as it is to dunk on yet another mediocre white man losing a fortune on a delusional fantasy, the lies our tech bros tell each other, and by extension tell you, are not without consequence. You might hear “$35 billion” and think, hey, that’s enough money to buy an NBA franchise or something, and you’d be wrong. It’s enough to buy five NBA franchises and have enough left over to buy a few MLB teams, and have enough left over to finance the next few Marvel Studios movies.

Subscribe now

The question the Aschenbrenner episode raises is the same one raised by so many similar episodes: How else could that time, that staggering amount of capital, and all of that very real, very useful human intelligence be more productively used? One is also left wondering about how much altruism was done by the Effective Altruist? Recall from Lies that a mere $1 billion provides a $10,000 per day allowance (before interest) for the 270 years, how much of the $45 billion did Aschenbrenner take off the table in order to make good on the “Give” part of EA’s Earn to Give”mantra? How many dollars worth of malaria nets are now deployed as a result of Aschenbrenner;s altruism? $40 billion? $4 billion? Or zero?

My guess is that it’s a lot closer to the last figure than the first two. Why? Because EA is a Silicon Valley cult that’s neither altruist nor effective.

I reached out to The Center for Effective Altruism’s media relations contact, but they did not respond to my request for comment.

Share

This piece draws on reporting, analysis, and excerpts from the book Lies My Tech Bro Told Me, which covers the real-world impact of Silicon Valley lies and what the non-billionaires can do about it.

For more on why space colonization functions as a distraction from solvable problems here on Earth, subscribe below.

And yet, I’m writing this for a reason. That reason is a recent Anderson Cooper 360 segment that boosted a carefully crafted marketing stunt from an AI doomsday cult known as Effective Altruism, which is one of the Es in the larger Silicon Valley quasi-religious belief system known as TESCREAL.

In any event, Cooper’s piece made me remember a segment on Google’s Deep Mind (like Anthropic, a frontier AI lab) reported by Scott Pelley, yet another respected 60 Minutes correspondent who recently resigned his own post after the Silicon Valley takeover of CBS News. The reason I thought of Pelley was this: whether we’re talking about him and 60 Minutes, or about Cooper Anderson (also a 60 Minutes correspondent), these reporters, and by extension their producers, reveal how utterly out of their depth they are when it comes to coverage of AI.

Here was Anderson’s segment, for those unfamiliar: A junior-level researcher named Jacob Coxon, 27 years old, quit his job at Anthropic, one of the big three US-based “frontier” Large Language Model (LLM) developers. (Google and OpenAI are the two others, followed by irrelevant players like Meta and xAI.) In the days leading up to Jacob’s resignation, he and a few colleagues worked on a shared Google Doc where they drafted and revised a series of social media posts that were to accompany Jacob’s notice. In other words, Jacob and his dimwit colleagues in the AI doomsday space (more on that in just a bit) were crafting his Anthropic departure in the same way a TikTok influencer crafts messaging about supplements or skincare. They were trying to get a post to go viral.

Coxon’s Tweet reads as follows:

Coordinating with several colleagues, the announcement was immediately retweeted in order to goose X’s visibility algorithm, as Coxon’s X account had barely been used prior to September 8th, 2026.

Even more revealing than the uncanny retweet timing, the Wall Street Journal had a story on Coxon’s dramatic resignation up on their website roughly 15 minutes before Coxon’s tweet went live. (The tweet went live at 5:04 PM Pacific, or 8:04 PM Eastern; the WSJ’s timestamp says 7:46 PM Eastern.) Consider the implications of that previous sentence. Unless Anthropic’s AI crapped out the instructions for time travel, front-running a tweet about humanity’s extinction with a press release to the WSJ seems a lot less like a troubled conscience wrestling with concerns over work they find morally objectionable, and rather more like evidence of a deliberate marketing campaign.

And yet. 150 million views and 200,000 new followers later, here we are.

Now, before I click ‘publish,’ is it worth investigating the initial re-tweeters who improbably stumbled on Coxon’s tweet within mere minutes of it going live? It is.

According to some sourced digging from Parker Thayer posted to X, those accounts included Nathan Calvin, General Counsel at Encode AI, Peter Wildeford, Head of Policy at the AI Policy Network, and Daniel Kokotajlo, Head of the AI Futures Project. Never heard of these dudes or these nonprofits, either? In what I’m sure is pure coincidence, a couple of web searches reveal that all three of these entities are AI doomer advocacy nonprofits. (AI doomer = AI will wake up and kill us.) Firms like the AI Futures Project, for example, predict that AI will take over the world in 2027. All three are middle-aged white men whose livelihoods depend on nonprofit fundraising efforts, and all three nonprofits raise funds based on the prophecy that AI is Skynet.

Speaking of the AI Futures Project and Daniel Kokotalo: is it worth mentioning where Daniel spent the afternoon on September 9th, the day after Jacob Coxon’s viral resignation tweet. It is.

He was on fucking Rogan!

So in a 24-hour span, we have Daniel Kokotajlo appearing on Joe Rogan for 2 hours, while Jacob Coxon appears on Cooper Anderson 360 for 10 minutes (before making the rounds on other outlets like Fox and MS Now.) As I said, I’m only a casual observer of Cooper’s work, but if there’s one thing I know for a fact, it’s that when it comes to a newsman’s reputation, Anderson Cooper does not want his name in the same sentence as Joe Rogan.

Moreover, when you pull a bit more at the thread of Coxon’s resignation tweet, you find the financial backing of the these nonprofits (like Kokotajlo’s AI Futures Project) from folks like Jann Tallinn (Skype billionaire) and Dustin Moskovitz (Facebook billionaire) and Max Nadeau of Coefficient Giving, yet another Moskovitz philanthropy. In fact, according to tax returns we have access to thanks to ProPublica, we learn that there have been direct cash payments from a Moskivitz firm to Jacob Coxon in 2022.

All of the men I just mentioned, and the connections between them, are explored in my book Lies My Tech Bro Told Me. In that book, I argue that what appears at a glance to be opposing viewpoints are, on closer inspection, not all that opposing after all. Whether we’re talking about an AI boomer like Altman, Musk, and Peter Thiel (AI boomer = AI will wake up and save us), or an AI doomer like Yudkowski, Moskovitz, and Jacob Coxon, all of these deeply strange Silicon Valley men are running the same grift. And that grift is based on a prophecy about Large Language Models which simply will not come to pass.

In fact, I can summarize this argument in two questions.

• If AI can do all you say it can do, why hasn’t it done that already?

• Conversely, if AI is going to kill everyone, why hasn’t it done that already? (And moreover, what are you doing to stop it?)

This is why The AI Doc was such a waste of everyone’s time, by the way! This is why Anderson Cooper’s CNN segment featuring Jacob Coxon was such a disservice to its audience. Just take an extra day before repeating every unverified and/or unfalsifiable claim! All of this stuff is trivially easy to track down - if I can do it, surely a $18 million/year newsman at 60 Minutes and CNN can do the same, or at least have a producer do some digging!

The reason why all this matters is because Anderson’s CNN audience, unfortunately, is other news outfits. For example, the audience is my local CBS affiliate who runs a headline about the Anthropic “whistleblower” in its crawl during its weather segment. (Coxon is in no way a “whistleblower,” a designation which grants specific legal standing.)

Even more unfortunately, that audience is policymakers who we’re counting on to understand the issue and respond appropriately. As of September 9th, Coxon’s tweet had replies/reposts from no less than 22 major politicians - 2 governors, 7 senators, and 13 house representatives. But Anderson’s piece helps precisely no one understand what’s going on. What’s going on is that we’re being conned, and we’re being stolen from, all to make a trivial number of white men, like Amodei and Altman, like Jacob Coxon and Daniel Kokotajlo, very very rich.

Even more depressing is that one of those senators was Bernie Sanders, who started his Twitter reply with “Mr. Coxon is right.”

Full disclosure: I’m generally a fan of Bernie. But he appears to be getting terrible, terrible advice from his staff about all things AI. (He supports a moratorium on data center construction for example, which I think is good, but not for the reasons given. He’s also suggested the government become shareholders in AI companies like OpenAI, which is a staggeringly bad idea for reasons we don’t have space to unpack here.) But as a final note (at least for now), I’ll add here that Bernie is right about Coxon being right about AI posing a grave threat to humanity.

The extinction level threat from AI is global warming.

It’s AI bros trying to cover the surface of the earth with data centers that consume massive amounts of fossil-fueled electricity. It’s data centers that produce massive amounts of heat as waste. It’s AI cloud giants like Microsoft, Amazon, and Google helping fossil fuel companies accelerate their extraction of yet more fossil fuels to be burned, to fuel the data centers, to produce the waste heat. And it’s especially the titanically stupid AI bros who think they can launch data centers into space from the moon.

Share

The threat from AI is that despite having all the knowledge and tools we need to begin reversing course on further warming the climate and setting off record El Niños in the Pacific, and glacier collapse in Nepal, and raging fires throughout Europe, the only problem AI seems really good at solving is how to generate sexualized images of Sonic the Hedgehog, or Donald Trump dressed up in a toga.

You got played, Anderson Cooper,1 and in turn helped countless other media outlets get played as well. I’m not saying you did it on purpose. But you’re getting paid too much money, and have too many producers available to assist, to be the unwitting mouthpiece of a strange, self-dealing AI cult.

I reached out to CNN’s prime time publicist, but she did not respond to a request for comment.

Predicting is fun!

I’ll add a caveat, however: Predictions, especially when made in public, turn out to be tough sledding. I say this from experience; I made a few in a book called Electric Cars for Dummies. As you might guess, those predictions involved which electric car companies flourish and which would fail in the coming years. I drafted that manuscript in the summer of 2022, the book was published that fall, and now the whole world has proof that I can be terrible at peering into the future. (I’ll save you the trouble of locating your copy: I predicted that Toyota would go out of business by 2030 because of their reluctance to electrification. But I also predicted that Nikola would go bankrupt, and that Tesla and BYD would thrive, so it wasn’t a complete showing of my ass.)

The point is that everyone occasionally engages in fortune-telling. Most times, it involves assessing available data to recognize patterns, anticipate risks, and prepare for what comes next. Especially when done in a team setting, it can spark discussion and generate new ideas, even when those ideas don’t pan out. Or, as Dwight Eisenhower famously said, “plans are worthless, but planning is everything.”

In fact, predicting the future is not only fun, but there are some who believe it’s a marker of intelligence. No. They believe it’s the marker for intelligence. As Elon Musk once told CNBC, “…the right metric for intelligence is probably the ability to predict the future.”

So by that “metric,” what might we conclude about the intelligence of the world’s richest human?

Or is this even a question worth investigating? After all, as Elon further said during an interview with Zanny Minton Beddoes, editor-in-chief at The Economist, “I’m very good at predicting the future. Not perfect, but my batting average is high for a human… I can see into the future not perfectly, but with a higher probability of success than most people.”

Well there you have it. Predicting the future is the key intelligence metric, and Elon Musk is very good at predicting. I guess there’s no point in looking into this any further.

As I argue here and elsewhere, the predictions being made and funded by Silicon Valley tech bros about “the future” - that is, about work, finance, immortality, and more - not only tend towards moral bankruptcy, but they’re often based on ideas borrowed from 100 year-old sci-fi. However, because these predictions are backed with billions of dollars, and because some humans will chase those dollars into whatever abyss those dollars are thrown, and because a good many of those humans are politicians who could rein in these harmful pursuits but choose not to, the tech bros predictions are ruining things for most everyone else.

One of the loudest and most laughably incorrect predictions the tech bros have trotted out for the past 15 years involves how we’ll soon be drinking red wine on the Red Planet, along with related pursuits.

As it turns out, smart people like astrophysicist Adam Becker, author of More Everything Forever, predict that human beings will not colonize Mars. Never. Ever. Or, as Defector writer Albert Burneko points out, there is no magnetosphere on Mars, and “any discussion of humans ever settling the red planet can stop right there, but of course it never does.”

For those interested in discussing beyond that - and I think it’s an interesting discussion! - the science behind why we won’t ever colonize Mars is well known, especially among those whose paychecks are tied to the pursuit. I’m thinking here about the founders of Mars One. This was a now-bankrupt Dutch company that “predicted” it would send its first batch of astronauts to Mars - a one-way trip, no less - by the year 2020. (To illustrate how gawpingly dumb that idea was, know that Mars One planned to fund the multi-billion-dollar mission by turning the entire process into a reality TV show.) I’m also thinking here about a company called GRU Space, which will happily cash your check to reserve a spot in their hotel on the Moon. GRU Space gets its financial backing from Silicon Valley’s startup incubator Y Combinator. I’m also thinking about the many, many employees of the publicly-traded company called Space Exploration Technologies. Those employees at SpaceX happily cash their paychecks despite knowing that we’ll never, ever colonize Mars. They do this because, as discussed elsewhere, humans suffer from “prognostic myopia.”

What’s that? The short version is that humans aren’t as smart as we might like to think. In particular, we’re not smart about avoiding harmful long-term consequences of our actions even when the harmful long-term consequences are all but certain. (I suppose this is where I insert the Upton Sinclair quote about it being difficult to get a man to understand something, when his salary depends upon his not understanding it.)

Of course, I’m especially thinking about Elon Musk, whose compensation as SpaceX CEO is tied to a permanent human settlement on Mars that supports at least a million residents. Musk receives one billion shares of SpaceX stock if that happens.

This will not fucking happen! But also, it doesn’t have to! He already owns over six billion SpaceX shares. The shares he already owns, combined with the value of the shares he owns in Tesla, are already worth roughly $1 trillion.

Why are these six billion shares worth roughly one trillion dollars and not worth, say, a few thousand, or even a million dollars? In large part it’s because the value of those shares is based on so, so many predictions. Yet despite what Musk claims about his batting average, the track record says otherwise. Almost all of these predictions are wrong.

In the book Lies My Tech Bro Told Me, I share what science has to say about Musk’s plans to colonize Mars with a million humans, and also what it says about launching satellites from the Moon using a railgun. In fact, I tie the two together. We know that the more immediate claims about lunar railguns are nonsense - Musk says they will be a thing within 10 years - precisely because of what Musk has predicted about setting up shop on Mars. Here’s a representative sample:

• In 2016, Musk promised to send unmanned missions to Mars in 2018, then manned missions in 2024. You may have noticed that SpaceX did not send an unmanned mission to Mars in 2018, and 2024 came and went without a single human anywhere near the Red Planet.

• In 2020, Musk promised to send manned missions to Mars in 2026. The timer on 2026 hasn’t yet expired as I write this, but one prediction I can make with 100% certainty is that there will be no mission to Mars in 2026, nor will there even be a serious attempt.

• In 2022, Musk promised to send humans to Mars in 2029. Still on the books as of this writing, technically, except for one small problem: SpaceX have now pivoted to the next item on this list.

• In 2026, Musk pushed the timeline for sending humans to Mars by a full two decades. We’ll now begin Mars colonization in 2046 thanks to the new priority, which is building giant railguns on the surface of the Moon, designed to fling AI satellites into orbit.

So that’s how we know there won’t be lunar railguns on the moon. Here’s what Musk actually wrote on the topic, per Leonard David at Space.com "By using an electromagnetic mass driver and lunar manufacturing," he wrote, "it is possible to put 500 to 1000 TW/year [terawatts per year] of AI satellites into deep space, meaningfully ascend the Kardashev scale and harness a non-trivial percentage of the sun’s power."

(I have plenty to say about dopes who mention the Kardashev scale in casual conversation, but I’ll have to deal with it in a separate post.)

While I was busy saving receipts from Musk’s predictions about Mars and the Moon, science reporter Sabrina Imbler, who also contributes to Defector, was doing the same. In a piece published in August of 2026, Imbler provides many more instances. (For the book, I figured 4-5 was a good “you get the idea” range.) What’s notable about Imbler’s coverage is how she takes the media to task for reporting/transcribing the Mars fabrications, rather than the man himself. In the setup for the article, Imbler asserts that: “Good science journalism is more interested in the truth behind big promises, or at least the likelihood of said promises coming true. By contrast, bad science journalism functions as a hype machine, regurgitating company statements verbatim.”

In any case, here’s what Sabrina calls a “small sampling” that communicates the “ubiquitous nature of this slop across a range of generally serious publications:”

• CBS News, 2016: “SpaceX Chief Elon Musk Predicts People On Mars In 9 Years”

• Los Angeles Times, 2016: “SpaceX could send people to Mars in 2024, Elon Musk says”

• Washington Post, 2016: “SpaceX says it will fly a spacecraft to Mars as soon as 2018”

• The Guardian, 2016: “SpaceX founder Elon Musk plans to get humans to Mars in six years”

• ABC, 2016: “SpaceX CEO Elon Musk Wants to Send Humans to Mars by 2025, Years Ahead of NASA”

• Futurism, 2016: “SpaceX Is Heading to Mars in 2018 and Sending Humans in 2025”

• CNET, 2016: “Elon Musk: We’ll have people on Mars by 2025”

• National Geographic, 2017: “Elon Musk: In Seven Years, SpaceX Could Land Humans on Mars”

• CBC, 2017: “This is how SpaceX will get humans to Mars by 2024”

• Wall Street Journal, 2017: “Elon Musk Aims to Land Humans on Mars by Middle of Next Decade.”

• Space.com, 2018: “A Base on Mars? It Could Happen by 2028, Elon Musk Says”

• Business Insider, 2018: “Elon Musk says SpaceX is on track to launch people to Mars within 6 years — here’s the full timeline of his plans to populate the red planet”

• CNBC Make It, 2019: “Elon Musk says his Mars base will have a ‘fun, outdoorsy atmosphere”

• CNBC Make It, 2021: “Elon Musk: ‘I’ll be surprised if we’re not landing on Mars within five years’”

• Business Insider, 2020: “Elon Musk says he plans to send 1 million people to Mars by 2050 by launching 3 Starship rockets every day and creating ‘a lot of jobs’ on the red planet.” (sic)

• New York Times, 2023: “Elon Musk Says SpaceX Could Land on Mars in 3 to 4 Years”

Yowser. And remember, that list is Imbler’s “small” sample. In the same article - here’s a gift link if you’re interested - they note that NBC carried a 2016 headline that read, “SpaceX CEO Elon Musk Says ‘We Plan to Launch People to Mars in 2024’”, which is the same article I cited in my bullet list of Musk’s failed Martian predictions. What’s more, NBC ran a separate headline earlier in 2016 that read: “Elon Musk Says SpaceX Will Send People to Mars by 2025.” (Neither of those two instances are included in the list above.)

So is this how it works? Rhetorical question. What I’m really asking instead, and what Imbler asks in her reporting, is essentially the same question, which is, How on earth does Elon Musk, whose companies famously do not employ PR departments, get most of the news media to fill this void? Do news outlets like NBC, The Washington Post, The New York Times, and many others, you know, look into the claims of someone in the news, especially when the person making those claims stands to benefit from credulous, uncritical acceptance of the claims, or super-duper especially if the claims, taken as factual and/or in good faith, might bring harm to the person acting on that claim? For example, if I claimed that adding cocaine to soda was a good idea for nursing mothers and their newborns, would NBC run that story? Or would that claim be fact-checked? I mean I’m just spitballing here, but if a prominent figure were to say something like drinking bleach is a great way to treat viral infections, shouldn’t the reporter hearing that claim see if the claim was rooted in scientific consensus? Isn’t that kinda the whole point of reporting?

When looking through Musk’s past predictions, we see that his “batting average” isn’t so great after all. By Musk’s own logic, then, we should conclude this is because he isn’t very intelligent. This is probably the correct conclusion! But also it’s not my main concern. Unlike much of Silicon Valley’s tech elite, I have no interest in the largely debunked notion of reducing intelligence to a single number.

The more pressing issue is that when those predictions arrive not in book or blog form, but instead via mass media and delivered to investors, such repeated, visionary claims get cast in a different light. I’m thinking here of the promise of 1000 new solar roof deployments per week by 2019. I’m thinking here of the promise of “over” one million robotaxis “for sure” by 2020. I’m thinking here of the promise of selling millions of humanoid Optimus robots by 2026.

These are just a handful of Musk’s predictions about companies he owns; whose engineers he can consult 24/7. But why limit yourself to predictions about fields about which you’re relatively well-informed? Musk sure doesn’t. I mean, who can forget this COVID prediction from March 2020, a topic about which he knows bupkis. On March 19th, Musk wrote, “Based on current trends, probably close to zero new cases in the US too by end of April.”

Lots of people in the media can forget, that’s who!

In any event, the issue is that when these predictions are made by the CEO of an aerospace company receiving large government contracts, they become outright lies. And the lies soon aggregate into a steaming mountain of bullshit.

So why does this matter? Why do I care, or why should you care, that Musk predicts, wrongly, that humans will land (have landed?) on Mars sometime in 2024? Or that a million humans will live on Mars sometime in the coming decades? Why should you care that all this Mars nonsense keeps getting uncritically reported by multiple otherwise prestigious news outlets? It’s their reputation, right? It’s his company and his money, right?

Wrong, or at least half wrong. It’s not your company, and you don’t have to read the news sources mentioned (most of whom generally do good work), but it is your money, and by extension your labor, that’s supporting this agenda. How so?

I’ll answer that in two ways. First, realize that even attempting Mars colonization will spell humanity’s doom unless something else does that far sooner. Thanks to companies that brought you single-use plastic silverware, and who profit from the incineration of fossil fuels, we’ve been trained to understand the concept of a carbon footprint, and I’m here to tell you that the carbon footprint of the first Egg McMuffin ever enjoyed on Mars would be the equivalent to hundred Exxons and Chevrons.

Second, this is a project that requires buy-in from almost all of humanity, yet almost none of humanity has been consulted about its merits. Recall that Musk’s SpaceX is a private company whose revenue stream has come almost entirely from US government contracts over the course of its 20-year history. And as of 2026, SpaceX is a publicly-traded company, one which was fast-tracked into indexes like the NASDAQ 100 after the exchange rewrote its rules in the days leading up to the SpaceX initial public offering (IPO). This means that workers saving for retirement all over the planet are passively funneling money into shares of SpaceX stock, six billion of which are owned by a single individual who can’t be voted out. In sum: it’s his company, but it’s your money (and mine) that’s bankrolling a morally reprehensible project, one which science tells us will ensure humanity’s demise.

The takeaway is that the bullshit is working. It’s time for that to end.

Share

This piece draws on reporting, analysis, and excerpts from the book Lies My Tech Bro Told Me, which covers the real-world impact of Silicon Valley lies and what the non-billionaires can do about it.

For more on why space colonization functions as a distraction from solvable problems here on Earth, subscribe below.

Subscribe now

In short, even the coolest product can’t be considered ‘done’ until people know how to use it, maintain it, and interact with it.

The phrase has a broader definition as well, referring to instances where specific written proofs, records, or forms are needed in order to validate some process. For instance, when you submit a claim for reimbursement to your employer for a client dinner, “documentation required” in this context means you need to supply the receipt as part of your claim, and no, you may not be reimbursed for your $500 bottle of Opus One.

In terms of naming this newsletter, this double meaning of the phrase is very much by design. It’s both a space that focuses on technology - especially who that tech is impacting - and is also a space that prioritizes receipts over hot takes.

Speaking of naming: the Quick Start part of documentation, as I’m sure you know dear reader, is the part of larger User Guide that helps users begin using a product right away. They’re the steps needed to unpack, assemble, log in, launch. Such is the case with this entry.

In the menu at the top of this site, choose Archive if you want to start at the beginning.

Otherwise, there are really three things you need to know for now:

• The newsletter is free to read. Sure, my family has to eat and pay rent and put kids through school like anyone else, but I want the topics written about here to be read widely more than I want to monetize content. I’ll figure something out.

• On a related note, if you’re interested in a beta read of Lies My Tech Bro Told Me, send an email here and put ‘reader’ in the subject line. Thanks in advance for joining the team to help launch the book; it’s perhaps the most impactful thing you can do.

• If you enjoy something you read here, the best way to say thanks (for now) is to comment and share. DR is not meant as a one-way street. The internet is a big room, and I’m hardly the smartest guy in it. So please, bring your brilliant insights and ideas into the conversation.

Oh, and I’m hiring a podcast producer. Paid position and everything. If you happen to be one or know one, shoot me an email at brian.culp@protonmail.com. Thanks!