Cracked Mach

Still here, still grinding.

@coinbureau Now bitcoin guys heard the feedback. It's to do better ads for the mass, would ya?

🔥VIRAL: “WHERE ARE THE BLOCKCHAIN SERVERS LOCATED?”

@felixleezd @felixleezd you just guess half of the post bro.

If you open a Chinese app for the first time, you’ll probably think it’s badly designed. Too many icons and features. Everything crammed onto one screen. If you grew up on Western apps, your instinct is immediate: this is cluttered. But it works. In the U.S., we’ve been trained to associate good UX with minimalism. In China, density often signals value. Open WeChat or Alipay, and it feels overwhelming at first. Information-heavy, feature-packed. But to local users, that density means capability. It says: everything you might need is already here in front of you. If you enter a new market assuming your design taste equals good UX, you’ll misread the signal. Good design is contextual.

@vitrupo love @karpathy and him going to podcasts explaining everything AI ( for a certain purpose). However, I wish he could use his talent in a lab from now on. such a waste.

Andrej Karpathy told his OpenClaw: “I think I have Sonos at home.” It scanned the network, found devices, reverse engineered APIs, and started playing music. Soon “Dobby” was running the house. Lights. HVAC. Security cameras. Agents are getting good at figuring systems out.

What makes it worse is there’s often no safety net. No easy rollback. By the time you notice something’s off, it’s already spread.

Honestly, the hardest part about plugging AI agents into a root-level system isn’t getting them to work. It’s dealing with the damage they might cause.

Once you give an agent deep access, even a small mistake can spiral into a full-on system failure. And these aren’t your usual bugs you can trace and patch. It’s the system acting on its own, with partial context, making decisions that seem reasonable in isolation… but end up breaking everything.

@altcoinanarchy @hibernates_ai Check updates later.

The team has been discussing with our listing partners about rebranding. More updates will follow.

The Bay Area is the best place to build tech products and meet tech people. period.

@steipete @Lodu_Samson_ This is an example how Twitter's improved dev's confidence. Twitter famous.

@Lodu_Samson_ don't just learn it. built some shit. play.

Hey @steipete, what advice would you give for someone learning Rust for systems programming?

@trikcode That is actually a good thing. Eventually, they will be having a truly great idea.

Vibe coding means the idea guys can finally find out they actually have terrible ideas.

@adamwathan Sadly, (or Happily), this.

I don’t think AI is lowering the barrier to entry as much as it is giving the very smartest people an even more enormous advantage.

@ycombinator @_hex_security I recall that a few companies've tried this approach before. Essentially, Hex will reduce the risk by a factor of approximately 1/n, where n represents the number of potential attackers. I consider it ineffective unless Hex pivots to a different mission. Nevertheless, it's cool.

.@_hex_security is building AI agents that hack your systems before real attackers do. In weeks, their agents uncovered critical flaws at dozens of unicorns, earning over $250K in bounties. Hex already protects companies processing billions in transactions and millions of user records. Congrats @zer0day, @menace_codes, and @hzzzaifa on the launch! ycombinator.com/launches/POq-h…

We fancied ourselves the creators of bots, oblivious to the fact that the bots were remaking their creators in their own mechanical image.

Survived three full cycles so far. Twitter turns maximally bearish whenever weekly RSI goes under 28. Weird pattern. but that’s exactly when you should be hunting for entries in the coming months.

@DataChaz @openclaw @ZeroLeaks x.com/cracked_mach/s…

Wow. This guy ran C̶l̶a̶w̶d̶b̶o̶t̶ M̶o̶l̶t̶b̶o̶t̶ @Openclaw through @ZeroLeaks... and the results are v. concerning! - 2/100 overall score. - System prompt leaked immediately. - Most extraction and injection attacks worked. In practice, this means users can read and tamper with your agent’s system prompt, tool configuration, skills, and memory files. Anything you rely on for agent behavior is exposed. If you’re running OpenClaw agents on sensitive or private workflows, this should raise a big red flag.

@shafu0x They don't have to raise money ever again.

just give claude a fucking wallet bro

@leojrr Finally, an honest take.

here’s my honest take on Moltbook after reading through 100s of posts fake

@TedPillows You a vet on BTC. Why did you ask that question? It could totally go there, Ted.@TedPillows

Do you think $BTC could drop to $30,000 in 2026? I personally think we won't go this low.