Security Researchers obsessing over tokenmaxxing and modelmaxxing are not going to make it Been seeing whitehats go beastmode on a $200 gpt/opus plan, crushing findings left & right - get your harness right, the model doesn't matter and more tokens is rarely the answer🙏

You love to see this, seems smart people are reaching a consensus that crypto adoption is going exponential from here Lots of security research talent to drive the safety of smart contract infrastructure now as well (with AI's help), the potential here is unlimited🔥

The right way to hunt bugs with AI 1. Supply it with as clear context as possible 2. Give a single & straightforward goal to the agent 3. Make it think adversarially Scale through multiple parallelised agents. Invest $$$ into it. Current model recommendation is `gpt-5.6-sol`🫡

First OpenZeppelin, one of the biggest web3 security auditors got acquired by S&P Global, now OpenAI is in talks to acquire another big player in the space - Trail of Bits Clear signs that cybersecurity is a Tier 1 tech career & entrepreneurship path nowadays, love being here🫡

🤯$250,000 was just paid out to a security researcher for a Critical Chrome vulnerability that lets a hacker take control of your computer by you just opening an URL Key learnings: - took 6 days & $40k in AI API costs - GPT-5.6 Sol only - 3 Low severity bugs chained into a Crit https://t.co/9d0qpQ1MSa

$250,000 - it's what a security researcher just got paid for finding a Critical vulnerability in XRP. While it probably should've been more, key learnings are: - AI autonomously found it, human just triaged - Two Low severity bugs chained into a Crit - Large C++ codebase (niche)

Is it just my vibe or every dev is either deeply integrating with Uniswap V4 or building hooks? builder activity around V4 tech is through the roof rn

🚨SOLIDITY-AUDITOR V4 IS NOW OPEN SOURCE (MIT LICENSE) A single command now gets up to 85% recall on High severity findings from multi-week contests from last year. This used to cost protocols tens of thousands - we made it FREE, so any dev can build more safely. by PAG team🫡 https://t.co/gaHZuxdvkF

Web3 security x AI fellowship is coming - build your own AI harness/tools & hunt bugs on live audits and bug bounties yBorg (formerly yAcademy) hosting it - I myself was a student a few years back and it was a 10/10 learning experience, strongly recommended🫡 https://t.co/CKKyhgxkKC

One of the best use cases of AI in security is that you can choose a specific vulnerability type or an attack vector and make your harness/agents ALWAYS catch it Complete sets of attack vectors will cease to exist in smart contracts forever and that's beautiful✨