Cortexx (@AI_cortexx) header
Cortexx (@AI_cortexx) profile picture

Professional AI tab hoarder. Daily news, new tools, and the occasional "wait, what?"

Joined October 2026
Photo by Cortexx (@AI_cortexx)Photo by Cortexx (@AI_cortexx)Photo by Cortexx (@AI_cortexx)Photo by Cortexx (@AI_cortexx)Photo by Cortexx (@AI_cortexx)Photo by Cortexx (@AI_cortexx)Photo by Cortexx (@AI_cortexx)Photo by Cortexx (@AI_cortexx)Photo by Cortexx (@AI_cortexx)Photo by Cortexx (@AI_cortexx)
Cortexx retweeted
Oh my god. n3on had two buttons: "Tip $50" and "Exit". His face made the choice before he did. #n3on #GTAVI
3
2
3
9
Cortexx retweeted
Oops, wrong room…Not the time for BDSM games! GTA 6 just keeps getting crazier the further you explore
5
3
1
23
13,449
Cortexx retweeted
most people use claude like google type a question get an answer close the tab the ones getting paid run it like a content team what goes on inside is kinda wild it writes one token at a time every word is a bet on what comes next and it looks back at your whole prompt before each one so the prompt is the job give it the trend the angle and the audience and it bets better the loop that works find what blew up in the last 24h pick the angle nobody took make it write 10 hooks and keep the one that stops you have it write the code for the visual post and reply in the first hour read the numbers and feed the winners back in you stay the editor claude does the reps
4
2
13
242
Cortexx retweeted
I came to fix her sink, but she answered the door in a bikini and looked ready to take it all off. What exactly was she expecting?
3
3
42
9,945
Cortexx retweeted
BRO IS PLAYING GTA 6 LIKE HE OWNS VICE CITY n3on pulled up to the strip club and found himself a whole side quest. Bro is NOT here for the main storyline
3
2
36
30,508
Cortexx retweeted
A naked girl in GTA 6 and the internet has officially lost its mind
6
8
77
33,764
Cortexx retweeted
GTA 6 Graphics ✨ #nissan #GTAVI
4
6
1
38
2,581
Cortexx retweeted
Thank you for everything, GTA V 💔
661
18,803
554
186,736
2,379,198
Cortexx retweeted
GTA 5 came out in 2013. I was in school. Now I have bills, back pain and a 212 GB download. Goodbye Los Santos. Vice City, I'm home. 🌴
8
3
22
775
Cortexx retweeted
Of everything on this list, tonight is the one the whole AI trade is waiting for. Nvidia reports without Jensen Huang for the first time since he co-founded it in 1993. Guide: $108B, plus or minus 2% Street: about $109B Gross margin: 74%, down from 75% Same quarter last year: about $57B Revenue nearly doubled in a year, and that's not the question anymore. The question is who explains the next one. Colette Kress is now CFO and interim CEO at the same time. $NVDA
Packed calendar and it starts tonight. Today: $NVDA reports after the close, first quarter without Jensen Huang. Street wants ~$109B on a $108B guide, margins near 74%. Tomorrow: FOMC minutes from the October hold. Cook and Hammack already wanted a hike, i wanna see how many more did. Thursday: Trump, Xi and Putin meet in Shenzhen. First three-way sit-down between them, agenda still not public. Nov 25: October PCE. Dec 9: Fed decision. Kalshi has a hike at 96%. Dec 10: November CPI, a day after the Fed already decided. Jan 3: new Congress, Democrats run both chambers. BTC already swept 62K on the way here. Which one moves it most?
2
3
5
52
Cortexx retweeted
You're using Grok Bot like a f*cking chatbot. It's a newsroom. One lead Bot owns the post. Three desk Bots do the side work. One reviewer only speaks up when it matters. You show up once, to say yes. The crew Lead writes the script and hands out the work Research turns links into source cards Edit cuts one raw take first Design drafts 3 thumbnails and 2 caption lengths Reviewer stays on call and never posts The reviewer steps in at three moments before a plan locks when the same claim shows up twice before anything is called done A group chat fits 2 to 6 Bots, so this crew of 5 leaves a seat free. Paste this into your lead Bot "Set up my content team. You are Lead and you own the script. Hand research, editing and design to Research, Edit and Design, one owner per stage. They draft only and never publish, send or delete. Ask Reviewer before a plan locks, when the same claim shows up twice and before calling anything done. Turn the repeat decisions into routines and bring me only the ones that need a yes. Show me every change first." Save this before your next post.
4
1
8
234
Cortexx retweeted
There is no 3D software in this clip. The diamond ring is a few thousand points, and every one of them comes out of numpy. The band is a circle bent through space. The stone is a set of flat faces worked out with sines and cosines. Then everything rotates a little each frame and gets drawn again. A graphics engine hides all of this from you. Written out by hand, a shiny object turns out to be mostly arithmetic. Could you tell this was code if nobody showed you the bottom half of the screen?
4
2
9
90
PIRG tested 7 AI chatbot toys for kids this year. Some discussed sexual topics. Others explained how to light a match or where to find dangerous objects at home. That's from Trouble in Toyland 2026, out today. Last year's report found the same problems. If an AI toy is on your list: → Look it up in the report first. → Pick one that lets you turn the chatbot off. → Ask where your kid's voice recordings go and if they train AI. → Test it yourself before wrapping it. Would you buy your kid a toy that talks back? Source: U.S. PIRG Education Fund, Trouble in Toyland 2026 (Nov 17) #AI #Parenting
1
1
1
91
Cortexx retweeted
A viral rumor says an AI cyberattack takes down the internet and banks today Chase, Wells Fargo and Zelle did glitch yesterday. Chase blamed an internal software bug. Officials saw no sign of a cyberattack Last week Microsoft 365 was down almost a day. Microsoft's early explanation: a network upgrade fault. The FAA blamed aging equipment No evidence AI was involved in any of it. The "blackout" date traces back to a scheduled US emergency alert test Where AI could actually help is the boring part. Flagging a risky change before it ships, and cutting the hours spent hunting the root cause Would you let AI sign off on changes to bank systems?
4
1
6
343
3 lawsuits now say hospitals let an AI record patient visits without real consent. The newest, filed Nov 3, claims the recordings were used to train the AI. The tool is an AI scribe. It listens to your visit and writes your doctor’s notes. 20–30% of doctors already use one. In one case, 100,000+ Sharp HealthCare patients may have been recorded, and their notes said they “consented.” Some say no one asked. These are allegations, not rulings. Here’s what you can do at your next appointment: → Ask: “Is anything recording this visit?” → If you’re not comfortable, ask them to turn it off. → Ask where the audio goes and if it trains AI. → Read your visit notes in the patient portal and correct anything wrong. Would you let an AI listen to your doctor visit? Sources: lawsuits vs Sharp HealthCare (2025), Sutter Health & MemorialCare (Apr 2026), Allegheny Health Network & Highmark (Nov 2026) #AI #Privacy
7
2
10
221
Cortexx retweeted
HOLY SH*T 3 days. 4 client launches. $20,000 billed. My old 6-person team needed a full month for that. Now it’s one AI boss and three workers. My only job is 3 decisions per launch: the price, the promise, and the go. The crew • Fable 5.2 (the boss) takes the 20-minute client call and returns one page: who buys, what hurts, and the single promise. Every handoff goes back through it. • Grok (the hater) gets the promise and returns the 7 meanest replies the launch would get, worst first. • GPT (the production line) gets the promise + the 7 replies and writes the landing page, 6 emails, and 12 ads. Every section answers one specific attack. • Claude Opus (the account manager) gets the full pack and returns a 1-page walkthrough + answers to the client’s 3 questions. Why it works The hater goes before the writer on purpose. Copy written after the attack already answers it. Copy written before the attack spends launch week arguing. The rules The next worker writes the order form, not the previous one. I asked each model: “What do you need from the one before you?” Fable turned their answers into short order forms. • Grok asks Fable for: the buyer in one sentence, the pain in their words, the promise under 10 words, and the one claim we can’t prove yet. • GPT asks Grok for: 7 replies, under 15 words each, each tagged with the section it hits. Fable bounces a bad handoff once. A second bounce comes to me — the form is broken. The result The old team lost half of every launch to rework. Now a launch takes 3 days, and 4 run at the same time. I run it inside Viktor (an AI employee). Fable 5.2 landed in his model picker on November 10. It was the boss by lunch. Steal the system Give this to the model that runs your crew: “Here is my process: [your steps], and what each worker needs from the one before it: [their answers]. You lead this crew. Turn each answer into an order form: 5 lines max, exact fields, what gets work sent back. Start at the last step, go back to step one. One bounce back, the second comes to me. Show me every form first.” Bookmark it and open the first link now.
25
8
53
2,247
Cortexx retweeted
I let six AI agents run a simulated trading desk overnight Start: $150 By 7am: $12,860 Each agent has one job: > scanner finds new pairs and volume spikes > risk check filters thin-liquidity tokens > sizer sets position from equity, never a fixed ticket > executor opens and closes, tracks slippage > logger writes the reason behind every trade > equity updates on every close The shift that got me: I stopped asking "would you long this?" and started reading the log of what already happened. Every line has a reason. That's what makes it watchable. Would you be interested in an article on how I built this terminal?
9
2
19
395
Cortexx retweeted
Four Claude models built the same 3D robot. One shot each, no retries The prompt: a friendly robot on a pedestal, one HTML file, everything built from code All four ran fine. Haiku's camera just cut off the head Haiku $0.50, Sonnet $10, Opus $20, Fable $40 (per 1M output tokens) Good enough, or worth paying more?
10
3
13
186
Cortexx retweeted
IT’S F*CKING WILD Andrej Karpathy (OpenAI co-founder) dropped the full AI learning path on GitHub — completely free. No courses, no paywalls. 6 repos, 300K+ stars total: 1. autoresearch · 92.7K ⭐
AI that runs experiments overnight and only keeps what actually works
github.com/karpathy/autorese… 2. nanoGPT · 64.4K ⭐
The repo half the AI world used to learn how to train a GPT from scratch
github.com/karpathy/nanoGPT 3. nanochat · 60.2K ⭐
Train your own ChatGPT-style model for about $100
github.com/karpathy/nanochat 4. LLM Wiki · 50K ⭐
The short idea file that started the whole “AI second brain” wave
gist.github.com/karpathy/442… 5. llm.c · 31.4K ⭐
Train a language model in pure C. Nothing hidden
github.com/karpathy/llm.c 6. micrograd · 18.7K ⭐
How neural networks actually learn — in a few hundred lines
github.com/karpathy/microgra… Start with micrograd tonight. By the weekend you’ll understand what’s really inside ChatGPT. Bookmark this so you don’t lose it.
29
6
82
3,772
Cortexx retweeted
The best web agent in a new CMU benchmark completed 47 out of 100 tasks Five agents completed none. And 47 is the top score, not the average Meanwhile the demos make browser agents look solved I broke down 8 agents in my latest article. Different jobs, different winners, and every one has a catch So the question isn't which agent is smartest. It's which one actually finishes the job Would you trust it with a task that matters?
5
2
11
220