The GB&U: August 28
The good, the bad and the ugly, from AI to autocracy, to vaccine and health fuckery, this is what caught my eye so far this week.

I am going to try to do this weekly, a survey of brief takes in the categories of the good, the bad and the ugly.
The Good
The Sandwich Man. The Hero we needed, a man in DC who got in the face of the goons who are patrolling DC and then committed a-salt by deli sandwich. The dude was toasty (beyond buzzed) and that had something to do. He offered to turn himself in, but no, Metro Police and the guards did a tour de force to arrest him at home with more than twenty officers going to his home.
Well, it turns out that a grand jury failed to indict for felonious sandwich tossing:

Cue up the Nelson Muntz meme:
Federal prosecutors on Tuesday were unable to persuade a grand jury to approve a felony indictment against a man who threw a sandwich at a federal agent on the streets of Washington this month, according to two people familiar with the matter.
I was in the middle of a meeting when this headline hit my phone, and I laughed loudly before I could mute. Awkward... Especially since I was on the phone with an ultra-MAGA colleague.
The rejection by grand jurors was particularly noteworthy given the attention paid to the case of the man who threw the sandwich, Sean C. Dunn. Video of the episode went viral on social media, senior officials talked about the case, and the administration posted footage of a large group of heavily armed law enforcement officers going to Mr. Dunn’s apartment.
Amen, some sanity by the citizenry.
On AI. If you've been here a while, you're probably aware that I am a product manager in tech, and have been doing this for, oh, 27 years now. Yikes.
Anyhow, what's blowing up the skirts of the movers and the shakers in Tech is AI, and in specific Generative AI. One thing that I have been is a skeptic. No, not that the technology is terrible and worthless, but more that it ain't AI, it doesn't "think" or "reason" and that while it is pretty good at some narrow tasks, it isn't the bee's knees.
As such, I read a gamut about it, both positive takes as well as the naysayers. One of the nay sayers that I am glued to is Ed Zitron. His post this week titled AI Bubble 2027 is just amazing. Alas a lot of the meat is behind the paywall, but he lays out why he thinks the AI bubble will burst, and how it will happen. It is not going to be a single solitary event, but a series of activities and events that will likely over the next 6 quarters lead to a significant alteration to the landscape.
And no, Generative AI will not disappear, but it will no longer be free, and it will become less pervasive in use cases that it frankly sucks at.
Last week, Sam Altman in the usual manner of his fishy talking mentioned that the current LLM based GPT AI may not be super extensible. This was a mealy mouthed way to say that buying more GPU's, and draining Great Lakes quantity of water to cool them, will not get to the holy grail, true General Artificial Intelligence. This caused more than one gasp, but quickly the media was back on the hype train.
In this post, Ed goes into details.
Setting the stage, Ed counters the fluffer's arguments from early in 2025:
Anyway, earlier in the year, a bunch of credulous oafs wrote an extremely long piece of fan fiction called "AI 2027," beguiling people like Kevin Roose with its "gloominess." Written with a deep seriousness and a lot of charts, AI 2027 makes massive leaps of logic, with its fans rationalizing taking it seriously by saying that the five authors "have the right credentials." In reality, AI 2027 is written to fool people that want to be fooled and scare people that are already scared, its tone consistently authoritative as it suggests that a self-learning agent is on the verge of waking up, a thing that is so remarkably stupid that anyone who took this seriously should be pantsed again and again.
I read that when it came out, and I was horrified at their poor reasoning.
He has a series of events that are likely to happen, but not in any order as this unravels (sorry for the long pull):
- NVIDIA's Growth Slows: As discussed in the Hater's Guide To The AI Bubble, NVIDIA is the weak point in the Magnificent Seven (which collectively account for 35% of the value of the US stock market) specifically because of its success (19% of the Magnificent Seven's value) and value, which is driven entirely by its ability to sell more and more GPUs every single quarter. It is inevitable that its growth slows, and once it does, the AI story crumbles with it. Perhaps it has a down quarter then an up quarter — but that's three long months to make the markets happy.
- I will add that there's a chance the bubble reinflates a touch if NVIDIA crushes earnings. Or maybe the market doesn't care? We'll find out at the end of next week.
- AI Funding Will Start Drying Up, And AI Companies Are The Most Funding-Dependent Companies of All Time: As The Information wrote last week, the "dry powder" (available capital to invest in totality) has dropped, and venture capitalist Jon Sakoda of Decibel Partners believes that "...if VCs keep investing at today's clip, the industry would run out of money in six quarters." Removing OpenAI and Anthropic, the party will continue until 2028...but we all know those are the companies that are getting the money.
- One of the major AI companies will collapse: OpenAI and Anthropic both burn billions of dollars a year, and have shown no interest in stopping doing so, with Altman doing his best Lord Farquaad impression and saying that OpenAI is "willing to run at a loss" as long as it takes to get...somewhere. For the bubble to burst, one of these companies has to die, and I will explain how this might happen.
- In both of these cases, going public will be nigh-on impossible, and even if successful will expose what I believe to be their rotten economics.
- Big Tech Will Turn On AI: Meta's AI hiring freeze isn't enough: Meta, Google, Amazon or Microsoft needs to bring an end to their capex burn, and they need to be definitive that it's A) happening and B) that it's happening, at least in part, because they have "built enough" or "exceeded the opportunity in AI." These companies are barely making $35 billion in revenue from AI in 2025, and I doubt that revenue is increasing.
- Another part of this: The Markets Make Big Tech Put Up Or Shut Up: At no point have the markets really interrogated the revenue from AI. I can imagine some weird fucking games happening here. Microsoft, the only member of the Magnificent Seven outside of NVIDIA willing to talk about AI revenue, stopped reporting AI revenue in January, when it was at $13 billion "annualized" (so a little over $1 billion in revenue a month). One has to wonder if this means revenue is flat or falling, as annualized is monthx12.
- I can also see a scenario where these companies start putting out obtuse "AI-enabled" revenue stats. I do not think this works, and if it does, it only puts off the inevitable, because they cannot avoid the capex crunch that's coming.
- This particular scenario is one that will happen in pieces.
- AI Startups Will Start Dying: As I've discussed previously, AI companies are currently raising at suicidal valuations that make it impossible for them to ever sell or go public.
- CoreWeave Will Die: AI data center developer CoreWeave is a time bomb, burdened with, to quote analyst Gil Luria of DA Davidson, "deteriorating operating income." It’s also a critical partner to OpenAI, providing compute as part of an $11.9 billion five-year-long contract, despite the fact that it's unclear if it has even come close to finishing its data center development in Denton, Texas. This has to be done by October, in part because that's when OpenAI is due to pay Coreweave, and in part because that's when it has to start paying back its massive, multi-billion dollar DDTL 2.0 loan.
- The Abilene, Texas "Stargate" project fails to get off the ground...or OpenAI can't afford to pay for it if it does. This alleged 4.5 Gigawatt data center will, when (if?) fully operational, allegedly lead OpenAI to pay Oracle $30 billion a year by 2028, which is more than OpenAI's combined venture capital and revenue to date. If this expansion doesn't happen it's bad, but even if it does, Oracle (unlike Microsoft) isn't going to accept "cloud credits," and for this to make sense, OpenAI would need to be making upwards of $4 billion in revenue a month and still have to raise a bunch of money.
I agree with this, and the rest of the article. If you care at all about the true state of the AI business, Ed is "must read" and worth the money for a subscription.

Second up, Gary Marcus, an expert in human cognition has one talking about the supposedly next level awesome image generation from Google, and while I won't quote too much from it, this image is too good not to share. AI image generation still has some issues:

So, that's what an asshole looks like...

The Bad
Let's start this with a post by Rick Wilson. Trump had flirted with rhetoric about being a "Dictator on Day one" during the campaign, but most of the right laughed that off as "Trump being Trump". Since his inauguration, it has been a steady stream of pushing the boundaries, and increasingly being authoritarian, all the while his enablers and the chucklefuck Republicans in the Congress have been laughing this shit show off.
It is no longer a stream, it is a fucking torrential deluge, and Trump is now claiming that "everyone" wants him to be a dictator.
Rick is careful to set up that he has been hesitant to call Trump a dictator:
He’s certainly dictator-curious, to borrow a phrase. He's nearing the end of his apprentice program at Dictator Technical School. His lust for it is seething, ugly, and constant. Now that he’s feeling the icy finger of death tapping his grotesque cankles, he’s pushing harder and harder for the Big Job.
He's wanted the scope and impact of dictatorial power for his entire life, the power to turn whim into law, to permanently turn the spotlight of all attention to himself, to finally be the richest man in any room, to crush enemies and reward friends.
Followed up with this:
The first step in any coup, soft or hard, is the collapse of the law. You don't need a single, dramatic act. You need a slow, insidious process. Both the Federal Courts, from SCOTUS down, have increasingly rubber-stamped Trump’s policies, accepting his absurd claims of unilateral power and immunity.
Yeah, we are in dictator land.
Read it all here:

Wednesday's Triad from the Bulwark is really good. The crowd at the Bulwark have been pretty consistently on the side that we will get through the Trump 2.0 and that some semblance of normalcy will return. I have vociferously doubted that, because the damage to our reputation on the world stage will not be fixed as long as other nations think that another Trump is just any presidential election away.
The internal damage is also going to be difficult to address. This Triad lays this out (usually, you can read the first part without a subscription):

The premise is that at this point in Trump's first term, his popularity was 38%, but now it is 44%. He is polling better than 8 years ago. That tells him (and me) that we as a nation are comfortable with authoritarianism, that the population likes what they are getting from Trump, and that going forward, they are in on an orange hued bully being at the top of the order.
This is fucking terrifying. It also means that we can't expect to elect Democrats and this will be fixed. There is a critical mass of hard support for all Trump is doing.
The third entry in the bad comes from former Bulwarker, Charlie Sykes, All the Marbles.
He goes into the insanity around yesterday's orgy of ass-kissing, also known as a Trump Cabinet Meeting, and it was bleak.
Perhaps you caught Tuesday’s marathonsnogfestcabinet meeting, where the president’s courtiers vied with one another in lavishing praise on the Orange Sun King.
Labor Secretary Lori Chavez‑DeRemer won the laurels when she effused:
“Mr. President, I invite you to see your big beautiful face on a banner in front of the Department of Labor because you are really the transformational president of the American worker," she said, while referencing a new banner unveiled on her department's headquarters.
It was a reminder that spittle doesn’t lick itself. And somewhere a North Korean minion blushes.
Recall that all these bootlickers were confirmed by the Republican led Senate (some with fucking Democrats also voting for them)
Worth the read:

The Ugly
I was seriously struggling with what to call ugly, but then like a bolt of lightning from above, I was reminded that Robert F. Kennedy jr. was still head of Health and Human Services.
Alas, the Bulwark has a pretty solid writer on all things health related (I assume that they hold their nose when Jonathan Cohn mentions that accessn to healthcare is a human right, but I digress).
Anyhow, Mr. Cohn points out that RFK jr could basically eliminate the entire market for vaccines in the US with one change, the protection that the manufacturers recieve to prevent them being sued by all these numbnuts antivaxxers.
In December 1984, one of the last two firms selling the pertussis vaccine announced it was getting out of the business, leaving the United States with just one. That news—along with warnings of coming shortages for other vaccines, including polio—got the attention of Congress. It responded by crafting a no-fault system designed to shield manufacturers from crushing liability costs while creating a quicker, smoother mechanism for compensating families whose children had suffered grave side effects from vaccinations.
The program is still in place today, alongside a counterpart program for vaccines like the COVID-19 shots that were developed quickly to address medical emergencies. These programs are the backbone of a liability system that is a big reason—probably the biggest one—that the vaccine market has stabilized, preventing shortages from becoming serious or routine.
This was originally around the Pertussus vaccine, aka "Whooping Cough". Still, it's protections are essential to keep the manufacturers doing business in the US.
And you fucking know that RFK jr is going to turn this knob to eleven.
Worth the read...

I do not have links for this, but I am seeing some whispering that RFK jr is planning on pulling approval for all mRNA vaccines, as early as mid September. This will mean that there will be no 2026 version of the Covid vaccine in the US.
It is already being delivered iin Canada, but so far, it hasn't been approved in the US, and if the fucking dickhead does this, there will be no Covid vaccines period.
Already, the latest variant is tearing through the country, and the fall isn't here yet.
Fuck.
(pre-pub update: the FDA is going to approve the 2026 variant, but with onerous restrictions. If you are not over 65, you are going to struggle to meet the requirements to get the vaccine. I do meet the requirements, but I expect to be grilled)
Garrett Graff has a banger of a post today, where he points out that we are currently operating at about two Watergate level scandals a week, every week. And we are becoming innured to it.
Almost every day there’s a scandal that just skates by that in any other moment of presidential history would launch endless follow-up stories, congressional investigations, and sink an administration.
It does get tiring. Just one of them is jaw dropping:
First, JD Vance and John Bolton. The search at the end of last week by the FBI of former Trump national security advisor John Bolton’s house was hardly surprising — Trump telegraphed throughout his reelection campaign that he was coming for his political enemies and opponents. And yet we’ve seen statements in the hours since the search that make clear that this process was probably even more corrupt than we imagine. Trump himself seems to walk right up to the line of bragging that he personally ordered the raid. JD Vance, in an interview, went even further: “We’re in the very early stages of an ongoing investigation into John Bolton ... if we think Ambassador Bolton committed a crime, of course eventually prosecutions will come ... there's a broad concern about Ambassador Bolton.”
It’s a stunning series of admissions, beginning with the use of “we.”
And then there's this:
Trump’s attempt to fire Fed governor Lisa Cook this week grew out of a very odd set of allegations of mortgage fraud that cropped up in recent days. The director of Federal Housing Finance Agency Bill Pulte has been on a one-man jihad in recent weeks against some invented scourge of mortgage fraud — he’s previously tried to launch criminal investigations of Sen. Adam Schiff and New York attorney general Letitia James — and then earlier this month posted on social media that Cook falsified information to get better bank loan terms.
CNN called it “the most consequential example yet of how Pulte has become an unlikely political attack dog.” The New York Times coverage, under the headline “Housing Official’s Push on Mortgage Fraud Gives Trump a Political Weapon,” was even worse — seeming to give Trump creativity points for weaponizing part of the federal government. As the Timeswrote, “Mr. Pulte, who has more than three million followers on social media, has used his position at the housing finance agency to attack and investigate Mr. Trump’s political enemies.”
This is the level of shit that got Nixon almost removed from office, and that was just by Tuesday of this week.
Fucking nuts. Read it all here:
https://www.doomsdayscenario.co/p/a-watergate-everyday
Well, that's all for this week.