Abhimanyu Gupta (Reverse engineering life)

be like me -> Do a shit ton of work somedays and earn money and then chill for some. -> Learning amazing theories of intelligence, psycology and philosophy. -> Do solid research in your field and write articles about it. -> Still have the ability to binge-watch the entire show in one night. -> Do sketching and creative shit which you can show off to people. -> Learn boxing cuz why not. -> Get amazing friends who all are doing well in their fields. -> Occasionally go on some dates. -> Smoke Sheesha. -> Try to be the best at every game you play until you are. -> Earn more money. -> Start your business on a whim and how much you can level up. -> Hate loud concerts and clubs. -> Love chillin in the lounges and cafes. -> Watch anime. -> Upgrade your hobby (Mine is sketching). -> Sometimes be strict with yourself and let loose most of the times. -> Be a hacker. -> Still listen to 2010s music. -> love the Japanese culture. -> Eat ramen. -> Have a home library. -> Automate your boring tasks. After all of this people still ask me if I get bored by myself.

there's a very clear point in loving somebody :)

Learning only happens by error correction.

🚨 Quantum Haber 🚨 siber güvenlik dünyasında çok ciddi bir adım atıldı. Tamamen otonom çalışan bir “AI ajan sürüsü” ile penetrasyon testi yapmak artık mümkün. Proje şu şekilde çalışıyor: - Recon (keşif) ajanı - Classification (sınıflandırma) ajanı - Exploitation (istismar) ajanı - Reporting (raporlama) ajanı Bu ajanlar ReAct reasoning ile birbirleriyle koordineli çalışıyor. Bug bounty, sürekli izleme ve CTF modlarını destekliyor. Go diliyle geliştirilmiş, Claude API’si ve 7’den fazla native güvenlik aracıyla güçlendirilmiş. Yani artık tek bir araç, profesyonel seviyede otonom penetrasyon testi yapabiliyor. 🔗 GitHub github.com/Armur-Ai/Pente…

I just finished reading Charles Hoskinson's book on ZK. It is a book about the construction of SNARKs and the current landscape. There is some mathematics in it - it is not always an easy read in some parts - but it's not a math book. It won't teach you how to build a SNARK beyond a high-level overview. I think the book makes more sense for those who already understand the topic to some extent, although anyone can read it. It's always good to get new perspectives on the subject. Another nice point is that it is quite recent and discusses recent advances in the field, more specifically lattice-based SNARKs and folding schemes. I found the book a bit repetitive in some parts, but overall it was an enjoyable and worthwhile read.

was having so much trouble during static file analysis so I made my own tool lmao; github.com/Hackerjedi666/… Check it out, just give this tool the binary you want analysed and it will tell you what type of file or malware it is Later on I will add the ability for it to detect if the file is actually a well know malware or not. Probably will use something virustotal api, nothing special just a small vibe coded tools, which i plan to make more in the future lmao.

My conversation with Sean Bowe (@ebfull) and Dev (@zkDragon) on the architecture of freedom. 0:00 - Why Dev’s working on Zcash 12:24 - AI is surveillance’s ultimate weapon 13:33 - Encrypted money must scale 15:49 - Compliant “privacy” is an oxymoron 19:25 - PIR for scaling Zcash 36:35 - How Tachyon and PIR work hand in hand 43:47 - Quantum recoverability 47:18 - Is quantum the last cryptography problem? 51:41 - Easier coinholder voting process 1:03:15 - Zcash roadmap 1:17:34 - What Zcash looks like at scale

ANTHROPIC PAYS ENGINEERS $750K PLUS TO UNDERSTAND LLMS AND STANFORD JUST TAUGHT MOST OF IT FOR FREE IN 2 HOURS

On April 13, 2026, a missing bounds check in @hyperbridge MMR proof verifier allowed forged proofs to pass. 1 billion DOT minted. Two attacks, combined with opportunistic withdrawals from drained pools, leading to $2.5M in losses according to Hyperbridge. rekt.news/hyperbridge-re…

I can quote @naval quotes all day and still can't tired out of it. "The gut is what decides, the mind is what rationalises it afterwards."

These days my llm rate limits were the only thing telling me to sleep. And then I biught clause max.

The most important line in the cypherpunk manifesto: "Cypherpunks write code." Hughes understood something that most activists don't: ideas without implementation are wishful thinking. You can write a hundred essays about privacy. You can give a thousand speeches. You can build a movement of millions who agree with you. And none of it matters if you don't build the tools. "We know that someone has to write software to defend privacy, and since we can't get privacy unless we all do, we're going to write it." This is the cypherpunk ethic: don't ask for permission. Don't wait for laws. Don't trust institutions to protect you. Build the thing that makes protection unnecessary. PGP didn't wait for governments to legalize encryption. Tor didn't wait for permission to anonymize traffic. Bitcoin didn't wait for regulators to approve peer-to-peer money. The code came first. This is what DarkFi inherits. We're not lobbying for privacy-friendly regulations. We're not hoping institutions will decide to respect our dignity. We're not waiting for the political climate to shift. We're writing code. Anonymous smart contracts. Zero-knowledge proofs. Encrypted coordination tools. The infrastructure of sovereignty. Consider the alternative: trusting that power will voluntarily limit itself. Trusting that the institutions vacuuming up your data will someday decide to stop. Trusting that the same systems prosecuting developers for writing open-source code will eventually protect your rights. The cypherpunks didn't trust. They built. Thirty years later, we're still building. That's the inheritance. That's the work. Cypherpunks write code.

True 90% of the AI startups today are completely hollow. They are just API wrappers desperately trying to find a use case, and they are going to bleed out on the operating table. The tech is definitely being forced into places it doesn't belong. But I feel like this isn't a recurrence of the NFT bubble. It’s a mutation of the Dot-Com bubble. During the Dot-Com crash, the market hallucinated insane valuations for companies that had no business existing. Pets.com died. The market crashed. But when the fever broke and the blood was cleaned up, the physical infrastructure the fiber optics, the servers, the broadband remained. The hype is a hallucination. The underlying shift in computational architecture is a hard, clinical fact.

The parallels between NFTs of 2021 and AI of 2026 are hard not to notice. - People brag about "token" spending - You buy in, and then get rugged (the model gets nerfed) - There's an almost religious need to "onboard" people to the new technology. Maybe I should remove the "almost" qualifier... - The tech has some legitimately good usecases, but it's also forced into usecases where every reasonable person knows it doesn't belong - Both booms depend trillion dollar capital injections (cov** stimulus and trillion-dollar AI spending) - Gatekeeping/whitelist/artificial demand gen for "the best opportunities" (models "too good to release") - People worried they'll be poor if they miss the "generational opportunity"

@saxenism Crazy ser, I loved this attack, checkout my articles too lol x.com/hackerjedi666/… x.com/hackerjedi666/…

The Drift hack was the most impactful in our industry in recent times, b/c it forced everyone out of their slumbers. The industry had to pause and notice the new realities of securing web3 protocol. Join me as we have a short discussion around it and the potential solutions.

@TheVixhal One of the craziest article i have read lmao.