these mega IPO's feel like a top of the cycle,
the AI trade tipping over.
first it was Gold,
Oil,
AI computer,
Biotechs,
IPO lunacy,
.... roll over or cool off before the final push higher
🚨BREAKING: Nvidia will pay you $1,000 a month to host a mini AI data center at your house.
It looks like a regular AC unit sitting in your yard. Nobody walking past would know what is inside.
Inside sits 16 Nvidia Blackwell GPUs and Dell servers running at full capacity.
A startup called Span builds and installs them. They are backed by Nvidia directly. The whole unit bolts onto your home and you get paid for the power and Wi-Fi you supply.
Some estimates put the monthly payout around $1,000. That is rent money just for hosting a box you never touch.
Span says the units deploy significantly faster and cheaper than traditional data centers. That is exactly why Nvidia is backing the suburban rollout instead of waiting for more commercial land.
The AI boom needed more compute. It found it in the suburbs.
The grid is being rebuilt one backyard at a time. Save this.
MCP is the new SEO
one one wants to log into your app when their slamming prompts from claude / claude code / codex / chatGPT / opencode / pi / or what ever AI harness their using.
apps exposed over MCP can connect into where the customer and user is,
i dont think a lot of the business / non techy folk tasked with delievering AI transformation projects realise that AI is intelligence on tap
its so good, that when implemented at scale the costs will go insane.
users ask similar thing on max giga brain models.
queries costing several $ to get a dashboard from a tool that already exists, with gas-lightning tone on the readyout
the real play with AI transformation is to run the models and charge per use, or position as a toll booth clipping $ from your customersas token flow moves through your systems
AI tooling is fantastic except for who ever pays the bills
Marc Andreessen admitted on Joe Rogan that AI is making people less efficient.
The guy who funds half the AI industry. On a podcast. Just casually dropping it.
Same week: Nvidia's VP said compute now costs more than his employees.
Microsoft canceled 100,000 Claude Code licenses because finance couldn't stomach the bill.
Uber burned $3.4 billion in AI budget by April.
And here's the detail everyone's glossing over:
Uber didn't just adopt AI. They gamified it. Internal leaderboards ranking teams by usage.
They made burning tokens a competition. A sport.
It worked. Adoption went from 32% to 84%. Engineers loved it. They used it for everything.
They stopped thinking about whether a task needed AI. They just used it. For everything. Always.
And that's when the budget died.
The tool was so good that people stopped being selective about when to use it.
And the moment you stop being selective, the cost goes exponential.
Because token-based pricing means every thoughtless query costs real money.
This is the part nobody wants to name:
AI doesn't have a cost problem. It has an addiction architecture.
Flat-rate software trained an entire generation to use tools without thinking about cost.
Now AI billing is per-use. But the habit of "just use it for everything" didn't update with the billing model.
Uber built a leaderboard that rewarded maximum consumption of a product billed per unit consumed.
Then acted surprised when the bill arrived.
Microsoft's engineers unanimously wanted to keep Claude Code. Finance killed it. The people using the tool said it was the best thing that ever happened to their workflow.
The people paying for the tool said they couldn't afford how much the users loved it.
We built something so useful that the only way to sustain it is to stop people from using it freely.
And that contradiction isn't a bug in the business model. It IS the business model.
It's how every AI company makes money: build something addictive, bill by consumption, and wait.
its that time of the year again, where we remember what could have been
laszlo didnt just buy two pizzas for 10k bitcoin:native in 2010
he also wrote the first #GPU mining code and dropped it on bitcointalk for everyone
turned cpu mining into something actually powerful and helped secure the network when it was still fragile
those 10k btc were worth like $40 back then
today theyre worth $770 million
the pizza gets all the memes but what he really did was push bitcoin from experiment to something that could survive
🚀
#BitcoinPizzaDay
Why Irish founders do well.
"The average Irish person is more likely to be a good hang."
@Willob argues that the relatively uncoddled Irish youth grow up to be more outgoing and higher agency adults.
@collision@patrickc do you agree?
what should have been done (and what you should do now):
- never store long-lived tokens or creds in repos, even private ones
- rotate pat keys and api tokens regularly
- audit your vscode extensions — remove anything not from trusted publishers
- use secret managers instead of .env files in any shared environment
supply chain attacks on dev tools are getting too common. this one just happened to be at Github
Github just confirmed a breach on their internal repos via a poisoned vscode extension. Team-pcp claims they grabbed ~4k repos and are trying to sell the data.
this is the exact supply chain risk everyone ignores until it hits the company that hosts everyone’s code.
attack vector was simple: employee installed a malicious extension. once inside, they pulled internal source and org files.
Github says customer repos weren’t touched (but how many have been done on the same exploit), but that’s not the real issue.
the real issue is whatever secrets, keys, and tokens were sitting in those internal repos,
im sure its rough day in the trenches rotating everything...