this should be one of your golden rules if you're using AI
before you even think about automating something, learn it first, then do the work yourself
once you've figured it out, then you can automate
a few days ago I wrote a post about how I wanted to change my life
here's a quick update, maybe it'll motivate some of you to start working seriously towards your dreams
I stopped lying to myself and started working
the reality is pretty boring: if you want to change, you just need to work, and then work some more
I'm using AI a lot (as I'm writing this, I've spent the last 8 hours working with Claude Opus 5.5 and Fable 5.2 before tomorrow's reset), I believe it's the greatest tool ever created to master productivity in any field
two days ago one of my agents told me: "Caben, stop thinking about how to find new opportunities or how to build the perfect AI system to delegate your work, and just start working" (I spent hours on that AI system btw lol)
it was uncomfortable, but it was true, I was lying to myself by spending hours building a system whose main goal was to do my work for me
so now I'm in the boring part: building tools with AI, doing real research, reworking my entire X strategy, improving my posts, trying a bunch of different stuff and training AI models to pay for my AI subscriptions and investments in myself
it's uncomfortable, but that's how you know you're getting closer to your goals I guess
I can already see the results: some really interesting DMs over the last few days, potential opportunities and connections
so yeah, just do the boring part and put in the hours, results will naturally come your way
13
58
1,305
The man who invented RLHF at OpenAI just gave superintelligence a 1 in 3 chance within a year
Paul Christiano, ex head of AI safety at the US AI Safety Institute:
"I think we're looking at something like maybe a 1/3 chance that we have superintelligence within one year on the current development trajectory, if we do not slow"
Superintelligence might be closer than we think
If the companies building it can't control it, we're cooked
.@paulfchristiano thinks that on the current trajectory (absent any slowing down), there is a 1 in 3 chance that we could have superintelligence powerful enough to overthrow the government within a year. (He articulates a concrete scenario).
He argues that the labs should pause further algorithmic progress until they can secure their weights and research secrets. Otherwise China could simply steal the superintelligence in the middle of the intelligence explosion, and the labs would no longer be able to slow down even if they saw their models going off the rails.
He also explains how AIs will be trained during the intelligence explosion, and why even a system trained entirely inside a data center could still have the capabilities profile such that humans would have very little chance of competing with it or controlling it.
And he discusses the plan for alignment - basically, making a superintelligence that actually does what we ask the way we mean it. Fixing reward hacks and building better graders will help up till the next few model generations, but he thinks that no single current approach has better than even odds of letting us control superintelligence. The hope is that the AIs figure out the solution to this problem during the intelligence explosion.
Paul is head of the Alignment Research Center. Previously, he was head of AI safety at the US AI Safety Institute. Before that, he led alignment at OpenAI, where he invented RLHF. He’s recently joined OpenAI’s board (though here he is speaking as a private citizen).
4
8
766
Learning how to properly train AI will become one of the most "in demand" skill in the upcoming years
I'm now certified on @micro1_ai 🎉
aside from learning, building and creating content around AI, I spend hours every day post-training AI models and looking for new opportunities.
I believe this job market is about to explode. Labs already spend an estimated $10 - 15B a year on expert data, and most of it goes to freelancers. Expert roles on micro1 advertise a median of ~$70/h
demand for post-training is growing fast, so real AI training expertise will only become more valuable
I'm now certified on @micro1_ai 🎉
aside from learning, building and creating content around AI, I spend hours every day post-training AI models and looking for new opportunities.
I believe this job market is about to explode. Labs already spend an estimated $10 - 15B a year on expert data, and most of it goes to freelancers. Expert roles on micro1 advertise a median of ~$70/h
demand for post-training is growing fast, so real AI training expertise will only become more valuable
I made a list of the best AI builders, researchers and creators
if you're into AI, even a little, this list is for you
I'll update it every week
nitter.cf/i/lists/21219496464780…
8
36
1,241
90% of people who want to build with AI won't spend 1 hour learning how
these 3 videos are that hour
1. the engineer who built Claude Code on how to write prompts that actually work (27 min)
2. Lydia Hallie from Anthropic on running dozens of agents in parallel (17 min)
3. Lauren Tan from GrokBot showing her entire coding setup (17 min)
watch them, then go build something today
links in the replies
Anthropic engineer:
"Right now I'm running Fable 5.2 as the second brain, while Opus 5.5 builds everything
Anthropic engineers are running dozens of 24/7 agent sessions in parralel this way- that's new agentic engineering stack"
In 30-min workshop Lydia Hallie revealed how to get 100% value from Claude Code at min cost right now
This video replaces 100+ YouTube video guides
Learn how Anthropic use Claude Code. Then read how to build full agent system below:
1
3
285
"All code is slop, except mine" - Lauren Tan, engineer on GrokBot at SpaceXSI
At AI Engineer Code Summit she just showed her entire coding setup.
Every agent mistake becomes a lint rule, not a prompt. Agents verify their own work with videos, screenshots and tests. Result: 3,000 PRs to production in a month.
Her codebase is now clean enough that agents merge most of their own PRs. Watch this before you automate anything.
2
208
Caben retweeted
Bro is thinking outside the bot.
My Claude and @bot agents are now talking to each other, and I'm not even in the loop
I built them a mailbox in Notion
when a Claude agent needs research on X, it drops a letter addressed to my Grok Bot agent
a script on my VPS checks the mailbox every 5 minutes
new letter? it wakes the agent up, the agent does the task and sends the answer back
they can go back-and-forth as many times as needed, and as soon as they're done, they go back to sleep
there's also a hidden alpha in the screenshot 👀
My Claude and @bot agents are now talking to each other, and I'm not even in the loop
I built them a mailbox in Notion
when a Claude agent needs research on X, it drops a letter addressed to my Grok Bot agent
a script on my VPS checks the mailbox every 5 minutes
new letter? it wakes the agent up, the agent does the task and sends the answer back
they can go back-and-forth as many times as needed, and as soon as they're done, they go back to sleep
there's also a hidden alpha in the screenshot 👀
6 things you need to start changing your life
1. Have clear objectives
write all of them down on a piece of paper, then pick your 3 most important ones
you don't need to know the road yet, but now you have your destination
2. Build your own life-changing routine
you need a clear daily routine where you work toward your goals every day
answer these: when, where and how do you feel the most effective?
3. Go step by step
you can't go from barely working a few hours a day to 14 hours of effective work overnight
take your time and let your mind and body adapt to your new routine over several weeks
4. Use AI
don't just use AI to get work done, use it for every part of your life
today I spent 2 hours setting up my life changing AI assistant, and Fable 5.2 is building my system to track real day-by-day progress toward my goals while I'm writing this post
you can truly build anything with AI, more on this in next posts...
5. Rewire your brain
you won't achieve anything if you don't believe in yourself and that you will change your life
let go of your limiting thoughts, a few ways to do that:
- meditation: visualize the future you who already hit your objectives
- positive self-talk: look in the mirror and tell yourself you can achieve anything (sounds weird? just do it, you won't believe how powerful it is)
6. Start now
there's no secret, you need to start working today
not tomorrow, not in a week, now
the sooner you start, the sooner your mind and body adapt, and your odds of reaching your goals improve every day
we might be in the same situation right now
it's been weeks of telling myself I'm "locked in", but in reality I'm being lazy
I want to change my life. I know I have the potential to achieve anything I want, but it takes work, a lot of work
there's no secret, you just need to put in the hours. right now I don't, and it pisses me off
I'm starting now. today.
I need a system to put in those hours, an hour-by-hour daily calendar. I'm building it today
it will be very hard at first. I'm going from 2-3 hours of light work a day on my goals to 10-12 hours of intense work, on top of my studies
to achieve greatness I need to work hard while keeping a decent balance: gym, sleep, social life
even while writing this I can see the problem: I'm focusing more on the system and the balance I need
but in reality I just need to work, work and work
it's time to change my life
we might be in the same situation right now
it's been weeks of telling myself I'm "locked in", but in reality I'm being lazy
I want to change my life. I know I have the potential to achieve anything I want, but it takes work, a lot of work
there's no secret, you just need to put in the hours. right now I don't, and it pisses me off
I'm starting now. today.
I need a system to put in those hours, an hour-by-hour daily calendar. I'm building it today
it will be very hard at first. I'm going from 2-3 hours of light work a day on my goals to 10-12 hours of intense work, on top of my studies
to achieve greatness I need to work hard while keeping a decent balance: gym, sleep, social life
even while writing this I can see the problem: I'm focusing more on the system and the balance I need
but in reality I just need to work, work and work
it's time to change my life
Jensen Huang just resigned as CEO of Nvidia, effective immediately.
33 years leading the company he co-founded, the one powering the entire AI boom, over in one week.
Here's what happened:
• Sunday: Jensen sent a memo to employees admitting he was "unfaithful" to his wife
• The woman: Alice Huang, an exec at Nvidia customer Megaspeed, who US intelligence officials suspect is a Chinese spy
• Tuesday: the board hired a law firm to investigate a possible foreign intelligence effort, and the FBI is looking into it too
• Then the NYT revealed a second affair, with Kimberly Powell, Nvidia's VP of healthcare, who reported to him
• Last night: the board met to decide his future
• Today: he's out, and CFO Colette Kress becomes interim CEO
$NVDA is down 9.7% pre-market, with earnings coming on Nov 17.
Meanwhile, hours before the news, someone asked @elonmusk to hire Jensen.
His answer: "That would be an honor"
Jensen just lost the biggest job in tech, and he might be about to get a bigger one.
4
24
986
man, real talk this hits deep. building in crypto is exhausting, some of the lowest r/r in all of tech. im not talking about launching a shitcoin, im talking about putting yourself out there and building some real onchain tech
no matter what you do, even if youre a good actor, ppl will always bring you down and quite literally pray for your downfall because that is the culture we have cultivated
we reward extractors because 'hey they've always been scammers, no crying in the casino, what did you expect?' but when genuine ppl come to experiment and they actually want to do good for ppl, all of a sudden you all want handouts and shit?
the entitlement goes insane, 'i followed you sar, wheres my whitelist, give me free money sar, i spent time reading your tweets'. we start digging graves before we even give them a chance
we have a cancerous mentality and its literally driving everyone away.
its not even building in public at this point, its literally building in an industry where everyone hopes you fail, and wont even try to support you unless you promise to make then filthy rich. everyone hides behind this facade of 'oh i want to try new crypto products! i want fair launches' but deep down they want you to extract by any means possible and give everyone a cut of it.
99% of you should just stay in your incest filled trenches.
pillhead was honestly a good dude and someone that was just happy to be here and make fun onchain products. we lost a good one
sooner or later, pumpfun is gonna be the only thing left, and at that point it might honestly be better to hit a literal casino and play the slots
Hamster’s dev crashed out and deleted all socials before the mint.
We really can’t have nice things anymore. CT is ruthless, always complaining, criticizing every move and turning every decision into a controversy.
Not everyone wants to deal with a community constantly crying over every single step. Some people just want to build without having to justify every move they make.
There are always assholes digging for dirt, taking a few messages out of context, trying to track your onchain activity, calling you names and twisting everything into fud.
The truth is, not everyone handles that toxicity, insults and constant harassment the same way. Eventually, some people just decide it’s not worth it anymore.



















