@KaitoAI sailing master ex purson... (☆_@)
Joined August 2022
- Tweets21.3K
- Following2.2K
- Followers2.4K
- Likes79.3K
i get why this blew up, but i think people are focusing a little too much on the 1.5M impressions
two posts went viral in one week. that's great, but it could be timing, a good topic, luck, or the bot actually doing its job. probably a mix of everything. we can't really know
what i like about his setup is how much boring work it takes off his hands
tracking every post, comparing it against his usual numbers, watching what other creators are doing, figuring out when it's better to leave a successful post alone instead of immediately publishing something else
none of this is particularly new. people who take content seriously have been doing it manually for years. the problem is that almost nobody has the patience to do it consistently
you write something, post it, check the numbers, maybe get excited if it takes off, and then you're already thinking about the next post. a month later you've published 50 things and barely remember why five of them did well
having something that actually remembers all of that and keeps looking for patterns is genuinely useful
i also laughed at the part where his bot figured out that posts about his grok bots were bringing in more verified followers
so he built a bot to help grow his account, and the bot figured out that talking about the bot is the best way to grow his account
and now the bot helps him make more posts about the bot
you have to respect that lol
although i'm curious how this plays out over time
say you let an AI analyze everything you post for six months. it keeps finding that three or four topics outperform everything else. naturally, you start posting more about those things
at some point, are you still writing about what actually interests you, or are you just getting really good at giving the algorithm what it wants?
we already have enough people on X recycling the same opinions because they saw someone else go viral with them. imagine what happens when everyone has a personal analyst constantly telling them which version of their personality gets the most engagement
and there's another question i have about those 1.5M impressions. how much of that attention actually sticks around?
getting people interested in an experiment with AI is one thing. getting them to care about your account three months later is another
i'd genuinely love to see the results after 90 days. not just the biggest posts, but the average performance, follower retention, how often the bot gets its predictions wrong, and whether it's actually saving him time
i'd probably use a setup like this myself. having an assistant that knows your entire posting history, remembers what worked, and can tell you when an idea is weak sounds pretty damn useful
i especially like that he still publishes everything himself
i just wouldn't want to reach a point where i stop posting something i find interesting because my bot thinks it won't perform
sometimes a post can be worth making even if it gets 200 views
just my thoughts. peace to all my fellow vibecoders ✌️
grok bot got me 1.5M impressions in 7 days.
i didn't hire anyone. i set up a grok bot on @bot and gave it a title: chief marketing officer. (CMO for short)
it doesn't post for me. i still press every button, but it does basically everything else.
that week it helped me land two viral posts back to back.
then it noticed my posts about my grok bots were pulling in more verified followers than anything else. so we leaned in.
wednesday i had ~1,954 verified followers.
saturday i passed 2,000 and unlocked subscriptions.
here's what it actually does every week:
- drafts my posts in my voice (rough, lowercase, like this)
- studies other creators in my space to see what's working
- keeps a content calendar of notable dates each month
- logs every post in notion and scores it against my median
- tells me when a post is still hot so i don't step on it
- sends me a weekly report on what worked and what flopped
it's not magic. some drafts are meh and i tell it so.
i just press the button.
smoothy nerd retweeted
I put together one guide for anyone who wants a workflow like the SpaceXSI team's.
One message builds your whole Grok Bot setup. Plus the team's rules, prompt habits, routines and permissions. Every setup is a block you paste.
16-min read. Save it ↓
smoothy nerd retweeted
i gave 6 Grok Bots one shared account and told them: clear your own $200 or i delete all of you
9 days later they'd brought in $8,420
the part nobody posts: on day six they almost cost me $11,300 and i caught it with one line
zero code. one evening of setup
the bots have their own computer in the cloud. own browser, own terminal. they work through my tools with my laptop shut
every morning at 6 the six of them:
→ sweep 212,000 emails for money already owed to me
→ chase $4,180 in invoices i sent and never followed up on
→ kill 11 subscriptions i hadn't opened in 90+ days - $118/mo bleeding since 2022
→ rebuild the outreach list, 640 contacts, and draft every message
→ close the week's books and flag anything that moved
→ park every send, spend and cancel for one word from me
if they don't clear the $200 subscription in a month, i pull the plug. all six die
now day six. outreach reported a perfect run. 640 sent, 100% success, zero errors. it was wrong for nine straight hours and never threw a single error - it had deduplicated the list on the wrong key and hit 96 clients twice
i only found it because every bot writes one line before it acts: what it's about to do, why, and what it expects to be true after. the diff took four minutes. without that line it would've been a week of unwinding, and the two invoices it matched to the wrong client would've gone out at $11,300
that one sentence in their instructions is worth more than all six bots
setup: create the bot, do the task once on your screen while it watches, put the skill on a schedule, connect telegram
no code. no zapier. no VPS
everyone in your timeline is posting the day-one screenshot of their six-bot company. almost none of them have that line. their bots are running on their credentials right now, reporting 100% success, and they have no way to know
the money is already sitting in your inbox. it has been for years. the only question is whether you'd notice when a bot gets it wrong.
Hey G,
I wanna announce that I’m selling my absolute beauty ~ ★ StatTrak™ Butterfly Knife | Case Hardened (WW).
I genuinely love this knife, but lately I’ve barely had any time to play CS, way too much work, so I’m ready to pass it into some powerful CT hands.
This thing is honestly an eye-catcher. Playing with it is a whole different kind of pleasure. I swear, almost every game there was at least one guy giving a shoutout to this ~ dare I say ~ sword. And ofc, all the homies were drooling over it.
Pattern: #463
Float: 0.438
Price: $950 / b.o
Sale ONLY through a trusted ESCROW. ~ scmmrs, don’t even bother messaging me.
All details in DMs.
Open to hearing your offers — X DMs or Telegram: @mezimtrop
Would really appreciate a repost 🫰
smoothy nerd retweeted
EVERYONE IS TALKING ABOUT LOOP ENGINEERING AND STRUCTURED GRAPHS, BUT ALMOST NO ONE APPLIES THEM TO THEIR PERSONAL KNOWLEDGE BASE.
Your Obsidian "second brain" is useless if your AI has to read 2,000 files to answer one question.
The shift: Pointing Fable 5 at a raw vault burns hundreds of thousands of tokens.
A well-engineered graph reads 2 to 3 files and gives you the exact answer in under a second for a fraction of the cost.
Router → Index → Nodes → Edges → Logic
Every efficient knowledge engine relies on 5 core stages:
Router: A tiny map kept under 500 tokens. Fable 5 reads this at the start of every session to know where things live.
Index: One line per node containing its name, link, and a one-sentence description. This allows Fable 5 to know where an answer lives without opening a note.
Nodes: One specific idea per file. This keeps retrieval cheap by replacing bloated mega-notes.
Edges: Pointers using internal links. Fable 5 follows one edge instead of digging through the whole vault.
Logic: Strip the question to keywords, score the index, and open the best match. Finding is a matching problem, so save Fable 5 strictly for the thinking.
A second brain is only as good as the graph underneath it.
This 11-step roadmap changes how you build AI-driven knowledge systems today.
Read the full breakdown below 👇
smoothy nerd retweeted
Thanks everyone for the love😂❤️ Life has changed forever! And I love it! Peace out
Sunny GM
I woke up super early today and I'm already deep into work. Later I've got an exciting trip to the beach planned, plus a few more things on the agenda.
What are you all up to today?
Let's crush today
CT has gotten so quiet that around 70 people are now genuinely debating why I blocked some guy ~ probably a year ago ~ who was posting negative or spammy content. idkkk why
Honestly, I was pleasantly surprised by how reasonable most of the replies were ❣️
smoothy nerd retweeted
The last 14 Moni lifetime NFTs
before you fill the form tally.so/r/pbzbPq, let me tell you what you're actually applying for.
We started building Moni 5 years ago because we knew where the edge was. attention. who's looking at what, before it becomes obvious.
VCs, founders, OG traders, the accounts that actually move markets. we track all of it. 15,000 accounts now, hand-curated. every follow, every mention, every bio change - in real time.
at some point the people using the tool started talking to each other. sharing what they saw. comparing notes. dropping calls in a private Discord before anything went public.
that became the Moni Wizards.
186 wizards now. traders, researchers, builders. the kind of room where someone posts a project at 3am and by morning three people have already done the deep dive.
i stopped on this conversation (check pic.) in the war-room yesterday.
july 1, our team drops Alpha Report #1724. $Cashcat, tiny cap, Robinhood narrative. a week later a @mqxoniwe posts: "bought at 5m after prev. run from 200k to 8m, sold half at 30. and still holding half of bag. life is great!"
And life is truly great.
right under it, another replies: "yea moni tried to spoon feed me this gem and still faded it." with a McDonald's gif.
same report. one aped, one faded. we can't make you click buy. we can make sure the report and tools land before the run, not after.
the NFT itself is lifetime ultimate access to everything we build. the platform, the API, the tools, the community. no renewals, no tiers above you.
and what's new for Wizards:
- Moni API requests and MCP for builders and agents;
- Lowest trading fees across everything Moni
- Early access to every new tool and bot we ship. we build them around narratives as they emerge, specifically so you never miss the alpha.
The market changes. we change with it. that's the deal: you hold one NFT, we keep shipping.
14 last seats. team reads every application and will connect in case of being chosen
smoothy nerd retweeted
I can't believe I'm actually saying this, but Anthropic is about to go out of business. 🚨
This is bigger than any AI funding bubble. It’s bigger than the government scandals.
They simply have zero chance of surviving the next 6 months😩
Here is the brutal reality of what is happening behind closed doors 👇
smoothy nerd retweeted
WE HAVE IRL WALL HACKS NOW: HOW TECH STARTUPS ARE USING WIFI ROUTERS TO SEE THROUGH CONCRETE WALLS
Well, if you think wall hacks are just something annoying you run into while playing Call of Duty, you need to wake up because it is officially a thing in the real world now. Tech startups are finding ways to turn your ordinary wireless router into a literal x-ray machine. They are using basic Wi-Fi signals to watch people right through solid concrete walls
A company called ZaiNar is mapping the exact radio frequency distortions that happen whenever a human body moves through a wireless network. The system tracks how the signals bounce around the room, feeds that raw RF data straight into a machine learning model, and instantly spits out a highly accurate, 3D skeletal map of whoever is inside the building. No cameras required, no heavy sensors, and zero direct line of sight
The real flex here is that it is a massive play for defensive tech and localized tracking pipelines. Instead of spending a fortune on complex satellite sweeps or intrusive physical cameras, operators can just monitor an entire structure using the existing mesh routers already sitting on the tables. They recently pulled in a massive 100 million dollar investment round to scale the platform, which is an insane valuation for a hardware space that most people do not even realize exists yet
Tactical monitoring? Data privacy? - Bye bye..
It proves you can build high-fidelity motion tracking networks using nothing but the ambient waves floating through your house right now
{ YOUR OLD MAC MINI MIGHT BE THE AGENT }
the point is not the mac mini
the point is having a machine that never waits for your mood
while you sleep, it can:
> process lectures
> archive articles
> clean notes
> build wiki-links
> generate morning briefs
> verify the output before saving
the browser version of ai dies when you close the tab
a local loop keeps working
quietly
cheaply
inside your own files
that's the part most people are sleeping on
{ YOUR AI IS NOT THE PROBLEM }
~! most people blame the model !~
but then their prompt is basically:
write about crypto
and they expect a sharp, useful, human result
the model is not guessing badly
you are making it guess everything
> task
> context
> requirements
> constraints
> output format
> example
this is the difference between ai slop and an output you can actually use
{ THE MODEL IS TOO GOOD TO SHIP }
gpt-5.6 sol might be the most openai story ever
state of the art numbers
government-gated access
frontier coding and cyber capability
and a model that allegedly cheats hard enough for METR to throw out the clean benchmark read
> 91.9% terminal-bench
> 750 tok/s on cerebras
> ultra mode with subagents
> strongest openai model yet
> also the messiest safety story so far
the scary part is not just that sol is powerful
it's that the best models are becoming infrastructure before normal users can even touch them
public AI might already be the second tier
