TL;DR: Jack Dorsey may have read that infamous Citrini report a little too closely. Yesterday, he announced that Block, the fintech company that owns Square, Cash App, and Tidal, is cutting 40% of its workforce (4,000 jobs) and tied the move directly to AI. Some industry analysts say it’s the biggest workforce reduction as a share of total employees in S&P 500 history, and depending on who you ask, it’s either a serious warning sign or a prime example of “AI-washing.”

What happened: Dorsey said Block is restructuring around “intelligence tools,” arguing that smaller, flatter teams can operate more effectively with AI embedded into workflows. The cuts reduce the company to fewer than 6,000 employees. Notably, this wasn’t framed as a downturn response. Executives emphasized that gross profit is still growing and described the decision as proactive rather than reactive. Dorsey said he’d rather make a decisive shift now than trim headcount slowly over time because the latter can be “destructive to morale.” Also, presumably for morale, he wore a hat that said LOVE during the town hall where he spoke about the layoffs.

Some analysts point out that Block did triple its headcount between 2019 and 2022 (a claim that Dorsey acknowledged), and Dorsey is famously the man who simply couldn’t make Twitter’s economics work. To some, this looks like an overhiring correction with a new explanation. Or, as one person put it, “At this point it doesn’t really matter if AI actually will work. Companies are going to use it as an excuse to cut bloat.”

What’s next: Dorsey, probably the least popular person at a Silicon Valley party this weekend, predicts “the majority of companies will reach the same conclusion” within a year. And some say he just gave CEOs cover to make bigger cuts in their organizations. (Here’s a look at how AI could broadly change corporate structures.) The larger implication: As AI tools improve, companies won’t just automate tasks. They’ll redesign around smaller teams from the outset. So far, Wall Street appears to be rewarding that logic; Block shares jumped at the announcement. In the meantime, the company has offered few details on how AI is replacing 4,000 roles.

Bottom line: Block’s move may be the clearest example yet of a profitable tech company making a massive reduction explicitly because of AI. And despite disagreements around whether Block grew too fast, too quickly, this may be the first true stress test of how far AI-driven austerity can go. —AC

A walking encyclopedia

Auto companies and their tech partners are betting AI chatbots belong on your car dashboard—ready to answer questions about your route or the next rest stop in a hands-free, natural back-and-forth conversation. But does that same idea make sense when you’re on foot or on a bike?

That’s what Google is banking on with its Gemini-powered navigation. As a carless pedestrian and cyclist, I gave it a try in a few different cities on both coasts.

My first impression? I found it hard to get over the awkwardness of speaking aloud to an AI on the street. “What neighborhood am I in?” I tried to discreetly mumble into my phone on a crowded Manhattan thoroughfare, as instructed by Google’s blog post.

It did unfailingly know the right answer, though. Its knowledge of local trivia was encyclopedic too; Gemini could tell me all about why Mayor Fiorello La Guardia established the Essex Market, for instance (to crack down on pushcart vendors). And it was well-versed in local stores and restaurants.

But when it came to adjusting routes based on these conversations, the app tended to struggle. At one point, it added 13 miles and a tour of a neighboring city to my walking route because it misunderstood a trail.

For me, I’d skip the hassle of trying to coax the AI voice into tweaking directions. The local fun facts can be entertaining, but I’d rather just google it later. Then again, I’m not the sort of person who’s inclined to ask Siri to do things for me in the first place. Read my full review here.

The Good: Gemini’s knowledge of local trivia tidbits and stores or restaurants is spot on, and it’s very in tune with your surrounding environs.

The Bad: It often struggled to turn the routes it describes aloud into actual directions within Google Maps. Maybe it’s personal preference, but the experience of talking aloud to an AI on the street is not for me.

Verdict: Noise —PK

Burger King is rolling out an AI chatbot that operates through employee headsets, the company announced yesterday. The OpenAI-powered assistant, called "Patty" (yes, really), will answer questions about recipes, flag broken equipment, and update digital menus when items run out. Standard operational stuff. But here's where it veers into an alternate dimension: Patty also listens to every drive-thru conversation and generates "friendliness scores" based on whether employees say "welcome to Burger King," "please," and "thank you."

Managers can query Patty to check how their location is performing in terms of friendliness. The company's chief digital officer insists it's just a "coaching tool," though he admits they're also "iterating" on capturing the tone of conversations. The system is already running in 500 locations and will roll out to all 7,000 US stores by the end of 2026.

The internet immediately clocked the dystopia, with many comparing it to Black Mirror's "Nosedive" episode—the one where people constantly rate each other's social interactions, which determines their status in society. One commenter captured the absurdity perfectly: "So Burger King can't make the ice cream machine work, but suddenly they've built Skynet for manners?"

Adding fuel to the fire: The average Burger King worker makes $13.65 an hour—below a living wage. "Instead of paying them better they spend millions on technology to make sure they stay as slaves making them say please and thank you. Wild," said one commenter. Another person put things even more plainly: "Welcome to Burger King, I love you." —SM

Chaos Brewing Meter: /5