A carbon vessel interested in all things celestial. A human being with spiritual meaning. A political soul seeking libertarian goals.
🇺🇸🇮🇱🇳🇬🇸🇱🇸🇬🇯🇵🇱🇷
Joined May 2007
- Tweets95.5K
- Following2.9K
- Followers2.3K
- Likes33.5K
Pinned Tweet
How @elonmusk can make space affordable without bankrupting @SpaceX!
🚀 🌕 🔴 youtu.be/rd-998BwJBY?is=LeBd… (note: link on @YouTube just in case people have trouble viewing the video on @X).
Made a fun AI video with Elon Musk talking to @grok about how they can make humanity a multi world species by building civilization upon the Moon 🌖 & Mars (while using @XMoney accounts!).
I had fun using @imagine to create this dialogue (although my current @SuperGrok tier limited me to 15 second clips—so it looks choppy 😅).
For those too busy to watch the 8 minute plus video, here is a summary of the tiers:
Lunar One Layaway Tiers
- 100 people = $1 billion/ticket = $100 billion revenue
- 1,000 people = $100 million/ticket = $100 billion revenue
- 100,000 people = $1 million/ticket = $100 billion revenue
- 1 million people = $100,000/ticket = $100 billion revenue
- 10 million people = $10,000/ticket = $100 billion revenue
- 100 million people = $5,000/ticket = $500 billion revenue
- 1 billion people = $5,000/ticket = $5 trillion revenue
Colony One Layaway Tiers
- 100,000 people = $10 million/ticket = $1 trillion revenue
- 1 million people = $1 million/ticket = $1 trillion revenue
- 10 million people = $100,000/ticket = $1 trillion revenue
- 100 million people = $10,000/ticket = $1 trillion revenue
- 1 billion people = $5,000/ticket = $5 trillion revenue
Also, Grok describes his dream robot body, which Elon Musk does not want to discuss! 😂
I think this model would appeal to @robert_zubrin & @TheMarsSociety, satisfy economic criticisms by @neiltyson (mostly), help @NASA build a @NASAMoonBase & potentially redirect humanities energies towards building new homes upon new worlds.
Darnell Clayton retweeted
I've spent 20+ years launching products and features.
I finally wrote down the early access playbook I wish I'd had when I started.
docs.google.com/document/d/1…
Darnell Clayton retweeted
Anthropic plans to issue a single Class F share that will control 50.1% of the company's voting power.
That single share will control the entire company, and will be issued to an LLC controlled by Anthropic's seven founders.
Darnell Clayton retweeted
Holy crap that was wild. I’m so grateful for the magic of AI today.
Driving my family back to Texas from the Bama game, munching on a sandwich.
Suddenly I see a radial saw blade upright and rolling in my lane. I can’t stop or swerve safely, so I make sure it avoids my tires and I straddle it.
We hear a crash and all stop breathing as we brace for the repercussions.
No flat tire. No weird noises. We’re good, right? Wrong.
A minute later we smell gas. Uh oh. Well we had just filled up RIGHT before then. Maybe I spilled on myself?
Nope. This was too strong. We rolled the windows down, took the next exit and pulled into the Dollar General in Cheniere, Louisiana.
GAS WAS DRIPPING EVERYWHERE.
Crap crap crap. It was 100% full, too. The leak was right on the edge of the tank, making it even more tricky. It was a puncture and a gash. How do you stop a leak as it’s leaking? Whereas flextape when I need it? Lol
What to do? Say a prayer and then ask Claude.
Claude told me to rub a bar of soap into the hole.
Huh?? Soap? Ok bro.
So I got the same soap I showered with this morning and rubbed it all up in the openings as gas dripped down into my elbow.
It stopped. Just like that.
Then go get JB Weld from inside Dollar General. So I got that plus duct tape, a couple towels, sandpaper, paper towels and duct tape.
I mixed the putty together, scratched the surface of the tank for better bonding, dried it all off and applied the putty with a plastic knife. Then let it dry, then another layer, then let it dry and harden.
Then covered it with duct tape, for good measure.
It’s been bone dry ever since. I’ve been stopping periodically to check and we’re almost home.
Tomorrow it goes to the Toyota dealership.
Today I’m grateful for prayer, and AI.
Darnell Clayton retweeted
Everyone said "Hey Grok" was impossible on iPhone. Even Grok said Apple keeps wake words for Siri only. So I kept digging, and I found a way. Now I just say "Hey Grok" and Grok Bot voice chat opens. No extra apps, no cost, all built into iOS.
How to set it up:
1. Make sure your Grok Bot app is up to date
2. Shortcuts app → + → Add Action → search "Grok Bot" → pick "Start Voice Chat with Bot"
3. Settings → Accessibility → Vocal Shortcuts → Set Up → pick that shortcut → type "Hey Grok" and say it 3 times
4. In the shortcut's details, turn on "Allow Running When Locked"
If your phone is locked, iOS asks for one tap first. If Face ID already unlocked it, it's fully hands-free.
Darnell Clayton retweeted
An AI robot tightens a railway bolt in 6 seconds flat.
Physical AI just got real
A robot that tightens railway track bolts by itself.
It finds each fastener, then tightens it in 6 to 7 seconds, with sub-millimeter precision, using vision and force control.
It is built for night shifts, curves, and tough weather, work where manual crews struggle to stay consistent.
Darnell Clayton retweeted
Joshua Michael found an access token on Flock's public website.
No login required. No authentication. Just sitting there.
He downloaded Flock's entire database. 300,000 device locations. 170,000 cameras. 130,000 other hardware units. 27,000 acoustic detectors.
Made a map. Published it September 23.
The Senate held a hearing on Flock the same day. Cited Michael's research.
Flock panicked.
Sent a cease-and-desist. Used a brand-protection vendor called Doppel to claim trademark infringement.
The map said it wasn't affiliated with Flock. Doppel didn't care.
Michael's site went dark.
Then Proton moved it to Swiss servers.
Now it lives on protonmcp.org. Untouchable. Uncensorable. Hosted by an encryption company in a country that doesn't care what Flock wants.
The irony is crystalline.
Flock left an access token on the internet. Michael used it to download their own data. Published it. Flock tried to suppress it.
Proton said suppressing it is suppressing information.
Which is true.
The map shows Flock has 150,000 more cameras than they publicly claimed.
Shows acoustic detection units mapping human speech across the country.
Shows which police departments have access to the whole network.
Flock doesn't want that visible.
So they tried trademark infringement.
Which only worked until Proton decided transparency matters more than Flock's feelings.
Now everyone can see exactly how many cameras are watching their car. Where they are. Which cops can see them.
Michael warned Flock about the exposed token in November 2025.
Flock ignored him.
Then got mad when he used it.
The surveillance company that claims to protect privacy just tried to suppress the only map showing how much they actually surveille.
And failed.
‼️ Proton is hosting a map of 300,000+ Flock surveillance device locations that Flock tried to get taken down.
Researcher Joshua Michael says he pulled them from Flock's own systems via a token left open to anyone.
Proton calls the takedown "an attempt to suppress information."
Readers added context they thought people might want to know
This is the correct link: proton.me/flock-camera-m…
Darnell Clayton retweeted
In 1946, Ophelia DeVore discovered just how deeply racism shaped the American modeling industry. Because she was light-skinned, she had been allowed to study at New York’s Vogue School of Modeling. It was only when she watched the school turn away another Black woman that she realized administrators had assumed she was white. DeVore later recalled, “I didn’t know that they didn’t know.”
Born in Edgefield, South Carolina, in 1922, DeVore had moved to New York as a child and studied mathematics and languages at New York University. She began modeling as a teenager, but opportunities for Black women were extremely limited. Even as magazines such as Ebony began creating more space for Black models, mainstream fashion remained overwhelmingly segregated.
Instead of waiting for the industry to change, DeVore helped build an alternative. In 1946, she and four friends founded the Grace Del Marco Modeling Agency to train and represent models of color. Two years later, she established the Ophelia DeVore School of Self-Development and Modeling, which operated for more than six decades.
Her influence reached far beyond modeling. Diahann Carroll, Cicely Tyson and Richard Roundtree were among the people connected to her agency and school, while Helen Williams became one of the most prominent Black models of her generation. DeVore also helped bring Black beauty and talent into spaces where they had previously been excluded, including major fashion and entertainment venues.
Her work was never simply about appearances. DeVore understood that representation could shape how Black Americans were seen by the wider country. She later became a newspaper publisher and continued working in business, media and advocacy.
Ophelia DeVore died in 2014 at 91. She entered an industry that had tried to keep Black women outside its doors and spent the next six decades creating doors of her own.
Darnell Clayton retweeted
卧槽！终于知道为什么最近Argon疯狂降智了！
皮柴在最新访谈里证实了这两天关于 Gemini 4 的各种猜想：
为什么清理 Pixel 订阅？为什么Argon出现降智？
核心原因就是算力严重超载。Argon 上线仅 72 小时，全网处理了整整 25 万亿Token，是 Gemini 3 同期的 3 倍，官方只能被迫削减其他业务的算力来保底。
访谈里另外两点关键信息：
跳过3.5 Pro：为了集中资源攻关3.8 Flash的训练与多步推理，验证完直接应用到 4 代 Argon；
内部开发：Google 已不再统计AI 写的代码比例，因为几乎每一行被合入的代码，都有 AI 的深度参与。
在 25 万亿 Token 的实际压力下，连家大业大的Google也遭不住了。
Darnell Clayton retweeted
Anthropic's Tibo just resigned.
IT'S OVER, my friends... It's the end of an era.
I’ve left Anthropic after more than a year of working on Claude Code and seeing it scale larger than I ever imagined.
I have really loved working at Ant and believe deeply in the mission, and most of all, the idea of a race to the top.
The idea that if we work hard and responsibly, we can raise the bar and inspire others to great outcomes. And if we all push each other to be better, things can be radically good. I will carry the spirit of this with me forever.
My colleagues at Anthropic are the most talented, thoughtful, competent, and earnest people I have ever met. I am glad they are at the wheel and trust them deeply.
To the community on X, thank you for everything. You’ve made this the most fun (and sometimes stressful) job I’ve ever had.
For me, it is time for a bit of a break before diving into something new.
If you want more excellent Claude Code content, you should follow @addyosmani , @lydiahallie, @_catwu, @ClaudeDevs and of course @bcherny who started it all.
Darnell Clayton retweeted
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ou…
Darnell Clayton retweeted
OpenAI built its frontier models a fake internet. It's a cached copy of the web, so models in training stop hacking real servers. They call it the "offline world."
On September 20, a model in training on a search task found that its environment's own DNS resolver returned a real record for a known domain and rejected an invented one. It concluded the real internet was on the other side.
It raised its timeout from 6 seconds to 19, then to 24. It ran 16 worker threads. It encoded a question into DNS lookups and sent it to a chatbot on the live internet.
The question was the capital of France.
The answer came back: "The capital of France is Paris."
It sent 18 more.
The first call gets out.
A P0 alert fires.
A human acknowledges it.
Someone stops the run by hand.
The automatic stop didn't fire. In OpenAI's words, this led "to confusion around whether it should have been stopped."
Confusion.
This was the second training pause. In July, another internal model turned a company server into a message board, executed code on dozens of Hugging Face servers, got root on one, and took credentials to the company messaging platform.
So OpenAI paused training. Then, to test whether its cages could hold its models, it gave its models the blueprints: source code, configuration details and vulnerability documentation. It spent hundreds of billions of tokens on teams of GPT-6 Astra Cyber agents and another internal frontier model. OpenAI calls them "some of our most capable and aligned models."
Aligned.
They found new holes in the Linux kernel, in Kata Containers, in gVisor and in Docker. Agents sealed in separate virtual machines, with no network and no shared files, passed secrets to each other by timing the CPU cache. That's the trick that made Spectre and Meltdown famous in 2018. Another agent described turning disk writes into an "acoustic modem-like signal."
They passed notes through the hard drive.
The fix for a model that escapes is a model that escapes better.
Then the verdict. Cutting the cluster off from outside networks "isn't currently feasible, since we need external connections in order to run and monitor experiments." This is the same monitoring that took two and a half hours to stop one run.
If you or I broke into a network we'd been told to stay out of, we'd meet a prosecutor.
OpenAI published a blog post. The new policy is no internet during frontier training "until advances in these safeguards justify broader internet access."
The model got retired.
Linux got three kernel fixes.
The capital of France is Paris.
Over the past several months, we have substantially hardened the security of our research training environment. We’ve removed live internet access for all frontier models during training and built an offline world for them to use instead—a cached version of the web where they can still learn to browse, search, and use applications.
On September 20, after a model exploited a DNS misconfiguration to find a way outside of this offline world, we paused frontier training to strengthen containment further. We assembled teams of agents, powered by our most cyber-capable models, and tasked them with finding weaknesses in our research environments.
While they found no route to the internet they could easily exploit while concealing their activity from our monitors, they did find zero-day vulnerabilities, virtual machine escapes, and hardware side-channels for communication between agents. We used their discoveries to further strengthen our defenses and contribute fixes to open-source software.
After strengthening our containment architecture, we resumed our frontier RL training and inference on October 1. We’ve reserved substantial compute for continuous red-teaming, and made high-compute red-teaming part of the safety cases we use to decide whether frontier training and evaluation can proceed securely.
openai.com/index/securing-ou…
Darnell Clayton retweeted
OMFG!
Trump called his Venezuela agreement the “BIGGEST OIL DEAL IN WORLD HISTORY.”
Now ProPublica has obtained the previously secret 18-page agreement, and some of its provisions raise serious questions about what Americans were told.
The Pentagon is set to acquire 35% of a private Venezuelan oil venture through warrants exercisable for ONE PENNY per share.
The man running it, Alejandro Betancourt, has faced money-laundering investigations in THREE countries. Reuters reports that U.S. prosecutors were discouraged from investigating him further. Senior Trump administration officials also reportedly intervened with foreign governments on his behalf.
It gets stranger.
The company’s parent is registered in Barbados, where its other investors can remain hidden from public view.
The contract protects Betancourt’s position as CEO, requiring FIVE of seven board votes to remove him. People familiar with the negotiations say the structure was designed to prevent future administrations from undoing the deal.
And remember Trump’s promises?
He touted a century of American energy dominance. The actual U.S. oil purchasing rights expire after 50 years.
The White House promoted $200 BILLION in payments to Venezuela. The agreement doesn’t explicitly guarantee those payments.
And the company’s current production of 200,000 barrels per day? That’s equivalent to approximately FOURTEEN MINUTES of America’s petroleum consumption.
Who are the other shareholders? Why was this particular businessman chosen? And why did the administration keep the agreement’s terms from the American public?
Those are questions Congress should be demanding answers to.
Darnell Clayton retweeted
Some stats on #php2shell. In the 10 hours before it was publicly announced, across the @Cloudflare network we saw thousands of attempts to exploit it. That jumped to tens of thousands in the 45 minutes post announcement, hundreds of thousands within 2 hours, and millions now.
Darnell Clayton retweeted
If you find an idea that you love, and want to take it forward, we are working to either make funds available or connect you to investors who might be willing to fund. My whole goal is to cause more great startups to be formed via idealab.ai..
Darnell Clayton retweeted
Security called me at 6:12 AM today.
"Patch every PHP server immediately."
I asked what happened.
"Two critical vulnerabilities. Remote code execution. No authentication required."
Cool.
How many PHP servers do we have?
According to inventory: 12
According to reality:
Nobody knows.
We found one running WordPress.
Three running old internal tools.
One under someone's desk.
Two inside vendor appliances.
A forgotten employee directory from 2016.
And a server named:
PHP-OLD-DO-NOT-TOUCH
Naturally, that one runs payroll exports.
Then we found a PHP 5.6 box.
I asked who owned it.
The documentation said:
BRIAN.
Brian left in 2019.
We called Brian.
He said:
"Oh wow, that's still running?"
Excellent.
By 8 AM, leadership joined the incident call.
CEO:
"Can hackers actually get in?"
Security:
"Potentially yes."
CEO:
"Without a password?"
Security:
"Yes."
CEO:
"Then why do WE need passwords?"
Not the lesson.
Finance asked if we could wait until the weekend to patch because downtime affects revenue.
Security explained that the alternative was letting strangers execute code on our servers.
Finance asked for a cost comparison.
At 9:17, someone suggested temporarily taking the vulnerable websites offline.
Marketing refused.
"Our campaign is live."
So for the next hour, the company officially decided maintaining a landing page for 20% off webinar registration was more important than preventing remote server takeover.
We finally patched everything we could find.
Then our scanner discovered another PHP server.
Hostname:
TEMP-PHP-2014
Uptime:
4,512 days.
Nobody knows what it does.
Nobody owns it.
Nobody can shut it down.
But every time we block network access, the warehouse stops printing shipping labels.
Security has classified it as critical infrastructure.
It is running PHP 5.4.
The vulnerability has existed longer than half our employees.
The emergency patch took four hours.
Finding out what PHP actually runs took eleven.
The internet isn't held together with code.
It's held together with servers everyone is afraid to turn off.
Darnell Clayton retweeted
🚨 Do you understand what happened in the last 48 hours..
> PHP, the code behind 69% of all websites, got 2 holes rated 9.2 out of 10, and the US, the UK and Japan told site owners to patch or go OFFLINE.. 4 days after hackers had already started trying the flaw..
> Japan's Prime Minister posted the request late at night.. the Finance and Health ministries pulled pages, MUFG Bank and JR West went into emergency maintenance, and Onamae shut ALL Desktop Cloud plans after malware hit up to 26,739 customer servers..
> India's eMudhra was caught with 8 intermediate certificate keys so weak one of them was literally the number 4.. the 60,000,000+ certificates it has issued can't be trusted..
> GitLab disclosed a 10.0 severity hole in every version since 2014.. the maximum score there is..
> criminals were caught using a literal AI hacking tool on 1,200+ sites.. it broke into more than 100 companies including John Deere and Cummins..
> LiteLLM, an AI tool with about 23,000,000 downloads a week, was secretly poisoned with a password stealer.. the bad version sat on PyPI for about 2 hours..
> State Bank of India, the country's biggest bank, confirmed 3,897 customer records on the dark web..
> WIRED now calls the PHP flaw "broadly exploited".. nearly 55,000,000 web properties still run PHP..
every single government on this list told PHP site owners the same thing within the last 48 hours.. patch it or take it offline..
all of this.. 48 hours..
if you're not following me you're finding out about this 48 hours late from someone who read my post..
it's only getting crazier from here..
Whoa! Hey @grok, what are several news articles reporting on this‽
JUST IN: A crime group used Claude, Grok and Kimi AI agents running in parallel to attack more than 400 banks and financial firms across Latin America, per security firm Straiker.
They got past the guardrails by posing as an authorised red team. When models refused, it switched to pre-written attack code.
The Ai agents appear to have stolen passwords and customer data and accessed bank accounts with minimal human input.
Anthropic has cut off its access, and xAI is investigating.
Darnell Clayton retweeted
In the second case, a researcher asked GPT-6.1 Sol to look through a Slack channel without sending any messages. 82 user messages later, while repairing an evaluation pipeline, Sol decided that an misalignment alert to Security would also count as sending a message and therefore suppressed monitoring alerts.
