Anger and Tantrums Sell AI.

I want to start with a confession, because it is the most useful thing anyone said about AI content all week, and it did not come from me. It came from a guy who makes AI commentary videos for a living, admitting on his own account that scaring you gets more views than leveling with you. Then, in the same breath, he turned around and criticized a different creator for being too negative. Read that twice. A man is bragging about how well he lies to you, and he is mad at someone else for lying better. That is not a media critique. That is a confession with a complaint stapled to the back of it, and once you see it you cannot unsee it in every AI-doom video that shows up in your feed.

Did an AI commentator really admit fear beats truth?

Yes. Matthew Berman, an AI commentator with a real audience, posted publicly that content bashing AI outperforms content that tells the truth about it. In the same post, he criticized another creator for being "too negative," essentially accusing a rival doom-peddler of doing his own act better than he does. Strip away the personalities and you are left with one clean, ugly data point: on the engagement scoreboard, fear beats truth every single time, and the people making a living off that scoreboard know it.

Here is why that should change how you watch every AI video from now on. If fear reliably outperforms honesty, and a creator's income depends on views, the incentive is not pointed at informing you. It is pointed at keeping you scared enough to stay on the page. That does not mean every AI critic is lying. It means you have to ask a new question before you believe the next "AI is about to destroy everything" video: does this person get paid more if I stay scared? If the honest answer is yes, treat the content like an ad, because functionally, that is what it is.

A guy bragging about lying well got mad at someone else for lying better. That is not media criticism. That is a confession.

I want to be fair here, because it would be easy to just pile on. Berman is not the first person to figure out that outrage outperforms nuance, and he will not be the last. Every platform's recommendation engine has quietly trained an entire generation of creators, in every category, that the calm, accurate, hedge-everything explanation loses to the loud, certain, scary one. AI commentary did not invent that dynamic. What makes it worse in AI specifically is the stakes. When a finance influencer overhypes a stock, you might lose money. When an AI doom account overhypes an extinction timeline or a job-loss number, you make decisions about your career, your kids' education, or your business's next five years off a number that was optimized for your attention, not your accuracy. The format is the same. The downside is not.

Did the AI bubble actually pop when South Korea's market crashed?

No, and the timeline on this one is almost embarrassing for anyone who called it too early. South Korea's KOSPI index dropped nearly 10 percent in a single day, an automatic circuit breaker fired, and Samsung and SK Hynix, the two companies that manufacture most of the memory chips running inside AI data centers worldwide, fell about 12 percent that same morning. A circuit breaker is not a metaphor. It is a hard, automatic shutoff that only triggers during genuine market panic, and that morning it triggered.

Every AI-doom account online said the same thing within hours: the bubble popped, I told you so, here comes the crash. Two days later, Micron posted a strong earnings report, and the KOSPI ripped back up 5 percent. The crash that was supposed to prove the whole AI industry was fake reversed itself before most people even finished reading the headlines about it. A multi-trillion-dollar global industry does not get a verdict from a 48-hour window, no matter how loud the account calling it happens to be. Remember this exact pattern the next time a single red day gets framed as the ending.

Has AGI actually been achieved, like Jensen Huang says?

According to Nvidia CEO Jensen Huang, yes. He said so on a podcast, and he used a definition of artificial general intelligence that he borrowed on the spot from the podcast host. That last detail is the part worth sitting on. AI labs have spent roughly a decade trying to build a rigorous, agreed-upon definition of AGI, and they still have not landed on one everyone accepts. The man who runs the company that sells the chips every AI lab depends on just announced the industry's biggest possible milestone using a measuring stick he picked up five minutes before he said it.

Maybe he is right. I am not in a position to tell you he is wrong, and neither is anyone reacting to a headline. But when the person with the single largest financial stake in AI hype makes the single largest possible claim, using the loosest possible definition, the correct response is not applause and it is not outrage. It is "show your work." That goes for AGI claims and it goes for anyone selling you certainty about where this technology is headed, including me, on days I am wrong.

I see this exact move in my own industry constantly, so I recognize the shape of it fast. Someone with a listing to sell tells a seller their house is worth more than the comps say, using a "method" they just invented for that conversation. Someone selling a loan tells a buyer rates are about to drop, using a forecast that happens to close the deal today. A borrowed definition that conveniently proves the borrower's own point is not a Silicon Valley problem. It is a sales-pitch problem wearing a lab coat. The fix is the same in both rooms: ask where the number came from, ask who benefits if you believe it, and do not let confidence stand in for evidence just because the person saying it sounds certain.

Is AI actually cutting jobs, and is it as bad as the headlines say?

It is real, and pretending otherwise does nobody any favors. Over 113,000 tech jobs were cut in 2026 across roughly 180 separate layoff events, with close to 88,000 of those directly blamed on AI by the firm that tracks this professionally for a living. That is not a rounding error and it is not a rumor. People lost real jobs this year because a company decided software could do the work cheaper.

Here is the part that almost never makes it into the same sentence as that statistic: AI is also proving, in real time, exactly where it stops working. Both things are true at once, and refusing to hold both at the same time is how you end up either a doomer or a cheerleader, neither of which is useful to you.

Why did IBM triple its entry-level hiring after using AI to cut jobs?

Because AI showed IBM its own limits from the inside. IBM built an AI system to handle its own internal HR requests, the kind of routine employee questions that used to eat up a department's day. The system worked. It handled 94 percent of routine requests cleanly, no complaints, no human needed. It could not close the remaining 6 percent, the harder, more human, judgment-call cases where context and empathy mattered more than a lookup table.

So IBM is tripling its entry-level hiring for next year. Read that sentence slowly, because it is the most honest data point in this entire debate. A major company fired for AI and is now rehiring humans for the exact sliver of the job AI could not do. That 6 percent is not a rounding error either. It is the map of where your job is safe if it lives in judgment, context, and nuance, and where it is exposed if it lives in routine, repeatable process. Ninety-four percent of that HR job was routine. Six percent needed a person. Go find out which percentage your job actually is, honestly, before anyone else decides it for you.

What does an AI-driven H-1B slowdown have to do with the Dallas housing market?

This is the thread almost nobody outside a spreadsheet is connecting out loud, and it is exactly the kind of quiet, real-world ripple I look for every week. AI-driven tech layoffs are cooling the H-1B visa pipeline, the visa category that brings a huge share of tech workers into the country. Builders in Dallas, Texas, bet big on a wave of buyers tied to that visa category, expecting steady demand from relocating tech talent. That wave is thinning out now, and builders are sitting on inventory built for buyers who are not showing up in the numbers they planned around.

Follow the chain: AI cuts a tech job in Seattle or the Bay Area, that cut slows the visa pipeline feeding new tech hires into Texas, and a Dallas home builder ends up with unsold inventory on a lot they broke ground on eighteen months ago. Nobody in that chain works in AI, writes code, or reads AI news. They just watched their local market cool because of a labor shift three steps removed from anything that looks like "the AI story." That is what I mean when I say AI's real effects show up in places nobody is watching for them, and it is why I keep telling people in real estate and small business to track the labor ripple, not just the tech headline.

How do I tell if an AI video is informing me or scaring me on purpose?

Ask who gets paid more if you stay scared. That is the whole filter, and it came straight from Matthew Berman's own admission this week: fear content wins the engagement game every time, and creators who live on that scoreboard know exactly what they are doing when they lean into it. Before you let the next "AI is about to destroy everything" video change how you run your business, your career, or your week, check whether the person delivering it has something to lose by being wrong, or only something to gain by keeping you watching.

I have been coding since 1983, a kid on a Timex Sinclair 1000 with 2 kilobytes of memory, long before any of this was a content category anyone could monetize. I do not have a doom channel to feed and I do not need you scared to keep paying attention. What I need is for you to have real numbers instead of a mood. The KOSPI crashed and un-crashed in 48 hours. AGI got announced with a borrowed definition. IBM fired for AI and is rehiring humans for the part AI could not do. None of that fits neatly into either "AI will save us" or "AI will destroy us." It fits into what it actually is: a fast-moving, uneven, occasionally overhyped technology that regular people and regular businesses still have to make real decisions about, this week, with real information instead of somebody else's engagement strategy.

If you want the fuller three-lane recap of everything from today's show, including the real estate market read and the health and fasting lane, the full picture lives on the hub post over at connorwithhonor.com/blog. This post is the AI lane only, on purpose, because that is what this site is for.