“U.S. payrolls rose 162,000 in August, much more than expected; unemployment rate at 4.1%:”
The U.S. economy added jobs at a brisk pace in August, reversing a summer slowdown in hiring, while the unemployment rate held steady. Nonfarm payrolls rose a seasonally adjusted 162,000 for the month while the unemployment rate, as expected, held steady at 4.1%, the Bureau of Labor Statistics reported Friday. Economists surveyed by Dow Jones had been looking for a payrolls increase of 53,000.
Does this mean the “AI-job apocalypse” is postponed again? It’s macroeconomics 101: if AI is an economic tailwind, history dictates it will create jobs indirectly through increased production and consumer surplus, not destroy jobs. There are a lot of jobs that cannot be automated, like anything involving interacting with people or manual labor. People confuse “AI automation” for “automation in general.” Or existing jobs can change roles. For example, font-end developers are repurposed to audit AI code and set up AI run-time environments.
AI can still make huge progress, while not actually replacing workers So far, despite rapid gains in AI, the awaited or predicted “mass unemployment” has yet to come. You can be optimistic about progress in AI while rejecting predictions of joblessness.
On a related note, this went viral:
This logic has never sat well with me, because public companies are ruthless about cutting costs. If these jobs were truly useless, they would not keep them. They serve some function, but it’s harder to articulate what that is. “Email jobs” are concerned with coordination and edge cases. This is very common in insurance or advertising, as each client or policy holder has unique circumstances. Email, hence the name, is the ideal medium of communication for clarifying these edge cases.
Good opinions and IQ (for a possible future post):
In the post, “Why successful people often have bad opinions online,” I argue having good opinions is highly g-loaded and difficult, even if it’s just propaganda. Propogandists historically have been quite smart. It’s as hard to craft a message that a lot of people find appealing, as it is to convey a technical concept to a small audience. The first requires finesse and deftness with language, whereas the latter requires assimilating a lot of complexity. For an opinion that seems uninteresting or banal, the actual skill lies in conveying it well, which is g-loaded unto itself.
Nike’s downfall:
Nike was founded by Phil Knight, a native of Oregon who was a runner. The company is named after the Greek goddess of victory. If it had focused on producing great running shoes rather than woke social objectives, this wouldn’t have happened. But it fell victim to Portland. https://t.co/v5R4gpGW9q
— FischerKing (@FischerKing64) September 6, 2026
If Nike focused only on running, its size would be much smaller. At only $19.5 billion as of 2025, running shoes are just not that big of a market compared to activewear in general. Nike still has a market cap of $59 billion even after falling 80%.
Wokeness is commonly blamed, but I don’t think it’s it. The fact “big tech” has done so well despite embracing wokeness , until doing a 180 pivot after Trump was reelected, torpedoes the popular “go woke go broke” slogan. Companies which censored Trump and his supporters pre-2022, especially during Covid for questioning the CDC or for questioning the outcome of the 2020 election, were lavished with shoutouts on Truth Social or Twitter, Whitehouse invitations of their CEOs, and direct investment in their companies.
Share prices of public tech companies, and private valuations of AI companies, surged post-2021 as the public forgot or stopped caring about those companies being woke, assuming they ever cared that much. So I don’t think wokeness can be blamed for Nike’s downfall, although it didn’t help either. Target embraced wokeness too and its stock price has recovered.
I think its more like: Apparel is a brutally competitive, low-margin business with scarcely any IP or moats, save for trademarks. Tech is the opposite: users are locked in by network effects, and it’s all protected by a labyrinthine legal and patent system, and high margins. With clothes, anyone can set up a online shop and outsource the production. And except for luxury brands, most people do not care if a certain shirt has better “tech” than any other shirt. Nike’s downfall was inevitable, similar to Lululemon, Adidas and Under Armour, all which met similar fates when consumer habits changed. Its time was coming.
Ziv’s AI roundups are really long and I typically only skim parts of it, but this part stood out:
Adam asks, how would you feel if you learned your doctor cheated to get into medical school? One answer is, that depends. Did he earn his degree once he made it in?
The students have a choice, but if you make cheating easy, that choice is grim. The solution is not to ask the students to just. People don’t just, any more than AIs will just if you mess up the training incentives. Calling for ‘ethics education’ and ‘building strong moral identities’ is the proposed solution here, and I am confident that will do little. Build a better game, change the incentives, or accept defeat.
A common argument you’ll hear is, “What if your doctor cheated?” The implication is that condoning AI use in school means accepting dishonesty more broadly, including in other professions where the stakes may be much higher. But people have a good general intuition at knowing which professions actually matter or not, or where cheating would mask incompetence or not. That is why no one is too worried that their doctor or pilot “cheated their way through school.” The explanation or response, is there is no take-home test or AI-equivalent for medical school, residency, or the MCAT (which is strictly proctored in-person and electronic devices are not allowed) , so it doesn’t really apply. All of these professions involve considerable in-person training.