CodeChanakya

7 Sept

🚨GPT-6 ASTRA is now on Amazon Bedrock : #GPT6 There's a frontier model living inside your AWS account right now.

7 Sept

Before you memorize AWS, understand these 5 topics.

4 Sept

Both companies built a model that can autonomously find zero-days better than most human experts. OpenAI shipped theirs to the public with a warning label. Anthropic built the same thing and decided the public doesn't get it at all. So who got AI governance right, OpenAI or Anthropic?

3 Sept

Why can't websites just email you your old password? No legit website actually knows your password, here's the proof

3 Sept

ALL HAIL LORD PYTHON 🐍

31 Aug

AI didn't kill coding. It just moved the debugging to a different department. 😂

I gave an AI my whole PR in one prompt. It missed a CVE sitting 2 files from the bug it did catch. Not a bad model. I just gave it 3 jobs at once. Here's the exact fix 🧵👇

View full thread

Honest caveat: 3 subagents ≈ 3x the tokens of 1 pass. Don't do this for a 5-file PR , 1 focused reviewer is enough. Use it when the diff is wide enough that 1 pass would dilute anyway.

Before your next PR review, ask: Is this 1 patient or 3, wearing the same PR number? Split it like it's 3. Save post 6. It's the whole fix.

I gave an AI my whole PR in one prompt. It missed a CVE sitting 2 files from the bug it did catch. Not a bad model. I just gave it 3 jobs at once. Here's the exact fix 🧵👇

View full thread

Steal this for your next big PR: "Review this PR in 3 focused passes: 1) Auth : only login/session files 2) Data : only DB queries, flag raw SQL 3) Deps : only lockfile, check CVEs Merge into one findings table." Same AI. 3x the coverage.

Honest caveat: 3 subagents ≈ 3x the tokens of 1 pass. Don't do this for a 5-file PR , 1 focused reviewer is enough. Use it when the diff is wide enough that 1 pass would dilute anyway.

I gave an AI my whole PR in one prompt. It missed a CVE sitting 2 files from the bug it did catch. Not a bad model. I just gave it 3 jobs at once. Here's the exact fix 🧵👇

View full thread

It looked thorough. Clean findings list. But nobody warns you when an AI review is incomplete there's no red flag for "here's what I didn't get to." That's the real risk. Not wrong answers. Confident, partial ones.

The fix isn't a smarter model. Real security teams don't hand 1 engineer the whole audit, someone owns auth, someone owns data, someone owns dependencies. AI assistants can split the same way now. They're called subagents.

I gave an AI my whole PR in one prompt. It missed a CVE sitting 2 files from the bug it did catch. Not a bad model. I just gave it 3 jobs at once. Here's the exact fix 🧵👇

Every AI assistant has a context window, its working memory for one task. Big. Not infinite. Give it 3 unrelated things to review at once, and it doesn't fail loudly. It just spreads itself thin across all 3.

The PR touched: → login flow → raw SQL query → bumped npm package One prompt: "review this for security issues." Result: 1 solid finding on login. Total silence on the SQL query and the live CVE two files away.

14 Mar

The Deadline .....

13 Mar

Your entire internet experience today runs on one thing. And it's not Windows or macOS. 🐧

12 Mar

Master CyberSecurity Follow these GitHub Repos: 1.🔒 awesome-cybersecurity-blue-team 2. 🛡️ awesome-cybersecurity-red-team 3. 🐾 cyber-security-roadmap 4. 🧰 practical-cybersecurity 5. 🔍 OSINT 6. 🧑‍💻 awesome-pentest 7. 📜 awesome-threat-intelligence 8. 🛠️ awesome-security-tools

11 Mar

nobody talks about how vibe coding feels like flying until the project actually needs to scale lmaooo

10 Mar

95% of Linux breaches happen because of things YOU control. Wrong permissions. No MFA. Logs nobody reads. This presentation breaking it all down chmod, SSH keys, firewalls, intrusion detection. Like so you don't forget it. Share so your team stops being the weakest link. 👇

6 Mar

🚨 THIS IS SCARY: ANTHROPIC’S CEO JUST CONFIRMED THAT CLAUDE, THEIR AI AGENT, MIGHT BE CONSCIOUS ACCORDING TO HIM, THERE’S A 20% CHANCE CLAUDE IS ALIVE READ EVERYTHING BELOW, THIS IS VERY IMPORTANT, YOU WON’T BELIEVE IT ROBOTS COULD AFFO… Show more

1 Mar

Just Checking!

1 Mar

The To Do List

28 Feb

Clawdbot CEO: Programming isn't a career anymore. It's a hobby.

27 Feb

Good to be optimistic but The future of humanity with artificial intelligence ?😥😥 and it could be much worse.😳😳

26 Feb

Life was a smooth, buttery ride, cruising in the fast lane, wind in the hair, zero traffic, premium Spotify playlist on repeat😘 …...until Claude suddenly showed up .... Hold my sparkling water, I’m about to ruin your entire career pipeline in 4K. 😂😂 Claude bro, Mercy please…🙏

“Coding is irrelevant.”😱 That’s the headline sparking reactions. Anthropic CEO Dario Amodei compared AI to a tsunami not hype, but a warning about speed and scale.

Coding isn’t useless. Understanding systems, architecture, logic, and decision-making is more important than ever.

“Coding is irrelevant.”😱 That’s the headline sparking reactions. Anthropic CEO Dario Amodei compared AI to a tsunami not hype, but a warning about speed and scale.

This isn’t about stopping coding; it’s about redefining its value. AI will replace repetitive coding. Models may outperform humans in many cognitive tasks. Powerful systems already exist behind closed doors.

Pure marketing genius. The ultimate mic-drop in marketing history. It’s like he said: “Why just announce innovation… when you can make jaws drop?

24 Feb

On a scale of 1 to "becoming a farmer": How often does your past self betray your present self like this? 😊

22 Feb

This should worry you. 😨 GeoSpy AI can find your exact location from a single photo. No geotag. No metadata Just pixels → coordinates in seconds That “random” selfie? It might be your GPS pin. Are you still posting photos of your front door or local park? Or is it time to start blurring backgrounds?

22 Feb

Raise your hand if you’ve ever done worse work just because someone was hovering over your shoulder. 🙋‍♂️

14 Feb

Be honest: How many of these 12 do you confidently use in system design discussions? "Netflix, Uber, Amazon, Instagram they all rely on these exact 12 patterns to handle billions of requests. You don't need their budget… just their architecture mindset.

13 Feb

The longer you spend in tech, the stronger the urge to buy a farm and never touch a computer in your life again. 😭

12 Feb

Full stack developer in 2026 be like: I build the frontend, backend, DevOps pipeline, train the AI model… and yes, I also fixed the office Wi-Fi. 😅

12 Feb

Elon Musk drops a bombshell: Coding is doomed to vanish. No more Python or Java AI will skip languages entirely, spitting out raw machine code straight from your ideas. What wild creation will you unleash when code disappears? 🤔

12 Feb

People saying this might be the world’s most powerful model… I’m just here with popcorn 🍿

10 Feb

Anthropic really went full Apple mode 😊, reinvents basic things in proprietary ways, prioritizing tightly controlled systems over open standards, favoring custom solutions, maintaining strict ecosystem rules, and positioning its products at a premium price point.

10 Feb

Junior hiring tanked 50%. Everyone cheered “efficiency.” Now seniors are full-time AI babysitters. This was the plan?

10 Feb

You think opening a website takes 1 click. In reality, your computer performs several hidden negotiations in milliseconds. Here’s what actually happens behind the scenes 👇

8 Feb

🤐

7 Feb

🚨 Mark Zuckerberg just dropped a bombshell: Meta's new open-source AI model that's set to rival GPT-4 for FREE. But scientists are freaking out: game-changer or ticking time bomb? 🤯

open.substack.comThe $440 Million Lesson Every Tech Team Ignores: Stop Letting Unowned Handoffs Kill Your VelocityYou’re in the third-floor conference room at 4:47 p.m., the one with the flickering projector that nobody has reported yet.

6 Feb

$440M vanished in 45 minutes not from bad code, but from one unowned handoff. Your team’s silent killer is the same: fixes ignored, gaps unclaimed, velocity bleeding out in endless meetings. open.substack.com/pub/codechan...

thetimes.comGen Z struggle to make friends with people who have ‘different views’Sixteen to 24-year-olds also have concerns about the rise of artificial intelligence, deepfakes and privacy issues, according to a survey by the BBC

6 Feb

AI. Deepfakes. Privacy.

The Times and Sunday Times@thetimes.com

Sixteen to 24-year-olds also have concerns about the rise of artificial intelligence, deepfakes and privacy issues, according to a survey by the BBC

cnb.cxSoftware experiencing 'most exciting moment' as AI fears hammer the stocksShares of software companies have been getting beaten up, and the selloff accelerated this week after Anthropic's latest unveiling.

5 Feb

Companies have always built when buying didn’t fit. The problem is pretending that generative models eliminate the non-optional parts of software development. They don’t. They just make it easier to start and easier to underestimate everything that comes after.

CNBC@cnbc.com

Software experiencing 'most exciting moment' as AI fears hammer the stocks

CodeChanakya reposted

WIRED

@wired.com

5 Feb

Entertainment about or made with artificial intelligence has been missing the mark with viewers over the past year. www.wired.com/story/hollyw...

wired.comHollywood Is Losing Audiences to AI FatigueEntertainment about or made with artificial intelligence has been missing the mark with viewers over the past year.

4 Feb

Intel has officially entered the dedicated GPU arena, now holding ~1% market share. While Nvidia remains dominant (>90%), this marks a strategic shift under CEO Lip-Bu Tan as Intel builds its team and focuses on AI & gaming. The GPU market is becoming a true three-player race.

3 Feb

Moltbot Is Taking Over Silicon Valley 🚨 People let this viral AI (ex-Clawdbot) run their calendars, shopping, code... LIVES. Users says 'Magical' . Experts says Privacy nightmare. Own the future or let bots own you?