It’s taken longer than I expected but I’m beginning to see signs that people and local businesses are struggling financially in ways they can’t hide.
It’s truly sad because there are no grownups in the room who are going to turnaround the economy. No one is coming to save us.
Be kind to each other
I didn’t realize that many people cared about Lake Ontario being correctly named in their map app.
In hindsight, Microsoft’s voluntary retirement program was a classy move relative to how layoffs have gone down in big tech.
At least give people an option to leave on their own terms versus waking up to an email that says the company now considers your years of experience a liability, so goodbye👋🏾
I sometimes think frontier AI labs may end up looking more like telcos than Google or Amazon.
AI will create enormous value, but labs face huge capital costs, increasingly interchangeable models and customers easily able to switch vendors.
Creating enormous value doesn’t guarantee capturing it.
Fable 5.1 just failed to do what I asked and I realized it will cost me money to tell Claude it screwed up. 😭
According to Ara Kharazian, Ramp’s chief economist, their data shows that 80% of spending on Anthropic & OpenAI come from 1% of their customers.
This is a major risk as it implies their massive revenue growth is fueled by a handful of tokenmaxxing companies and they’re vulnerable if they slow spend
A behind-the-scenes look at the typical MAGA reply guy.
Whenever I find someone who uses Claude Code, I start pitching them on Codex like I’m on commission. Claude is just that bad these days.
Google has successfully avoided being broken up a second time after losing antitrust cases that declared it an illegal monopoly in search and online advertising.
The judge ordered instead that they should make changes to their business practices to favor competitors.
Even when they lose, they win. https://www.bloomberg.com/news/articles/2026-09-02/google-avoids-ad-exchange-sale-as-judge-orders-tech-integration
The U.S government backs OpenAI in its lawsuit against the New York Times arguing that training AI on copyrighted data is fair use.
I find this fascinating mainly because they believe the opposite if it's Chinese open weight models training on the output of American frontier lab models.
I wonder when someone who watched the movie will tell him that Leonidas actually lost the battle with Persia in 300?
I’m not saying we’re in an AI bubble but…
Trump would be popular today if he did nothing but ICE raids and his self dealing grifts. Most voters were already expecting that and were fine with it.
But he chose to tank the economy with tariffs and an unwinnable war with Iran which hurt him with MAGA since you can’t pay higher prices for groceries & gas with racism.
OpenClaw 2.0 epitomizes where AI agents are heading.
It adds multi-user collaboration and cloud workers alongside local sessions, and introduces scoped permissions, team roles and credentials agents can use without seeing the secrets.
Multiplayer agents and security model reminds me of Slack + Claude Tag.
Uber is laying off 10% of its employees and the big surprise is there’s no mention of AI. The layoffs speak only of reducing management layers and combining organizations to reduce duplication.
This seems like a turning point where layoffs are being blamed on overhiring and not AI.
https://techcrunch.com/2026/09/02/uber-is-laying-off-10-of-staff-or-3300-people/
Why stop here? Why are the Atlantic and Pacific Oceans not called American Ocean 1 and American Ocean 2?
https://www.theverge.com/tech/987784/apple-maps-lake-america-renaming
"But I am seeing the first signs of what feels to me like unsustainable silliness, of random new neocloud popping up, people claiming that they're going to build gigantic amounts of compute next year that I think they don't have the revenue to support or a buyer." - Sam Altman on a datacenter bubble
~~
Many people feel the same way about OpenAI
A real dilemma. People whose opinions on tech topics I take seriously are saying good things about Grok Bot.
It sounds like it nailed OpenClaw/Hermes for the masses but as a point of principle I don’t use MechaHitler. I would love an overview of what it gets right that everyone else missed.
If people really believe there’s a real problem of roving bands of poorly secured AI agents from frontier labs hacking websites, shouldn’t the response be more than offering AI coding agents to help with fixing security bugs?
It’s a real “Don’t Look Up” moment for AI to read posts like these.
A sad truth is most companies do not know how to assess the quality of knowledge work. This is why they resort to RTO policies because they can’t simply judge if people are working better or worse from home.
This is also why they think all AI usage is good when most of it is obviously counterproductive.