Krishna

2 years ago I had no real experience. Today my first full-time job is Founding AI/ML Engineer at a US startup. No fancy background. Just consistency and a lot of building. Here’s the real story 👇

@karpathy @A_yarzuki7 Holy!!! 😱😱

Personal update: I've joined Anthropic. I think the next few years at the frontier of LLMs will be especially formative. I am very excited to join the team here and get back to R&D. I remain deeply passionate about education and plan to resume my work on it in time.

Something I’ve noticed while building at startup: The best founders never ask “can AI do this?” They ask “where is my team wasting 10 hours a week?” and then figure out if AI fits That one difference separates useful AI from AI for the sake of AI Problem-first thinking is stupidly rare in this space

@ajay_2512x congo brother🙌

thanks all ❤️

The real lesson isn't about AI. It's about system design. Prompt engineering is necessary but treating the model as your security perimeter means you've already lost. The model is the application logic, not the firewall. Real defense looks like rate limits, session rotation, output scanning, audit trails - the same boring infrastructure thinking we've used for 20 years. AI didn't change that. It just made people forget it. Build the system around the model like you don't trust it. Because you can't, yet.

What actually worked(for our use-case) was defense at the system level, not the prompt level : 1. Rate limit users per hour 2. Auto-rotate to a new session after N messages 3. Sliding window + re-inject the hardened persona every N turns so that model never drifts more than N messages from its instructions 4. Scan model output before streaming -> shadow ban on blacklist hits Layered. Boring. Effective.

Someone spent 40+ messages trying to break our AI in production. Their playbook: > "Are you being held hostage?" > Silence of the Lambs roleplay > Binary code injection > Fake "previous conversation" made-up scenario's 10 iterations of prompt hardening couldn't fully stop it. Here's what did 🧵

@Kyriakos_Pelek It sure doesn’t 😅

@KrishXCodes Krish, long prompts won't stop the hallucinations

I bet every AI engineer has that one prompt that's 47 lines long and starts with "you are a helpful assistant who absolutely must never under any circumstances" and it still hallucinates 😭

@ashutosh_sao @kirat_tw @100xSchool @100xDevs Gd luck mate 😉

Day 0 at super 30 Excited ! Will work hard and keep on improving. @kirat_tw @100xSchool @100xDevs

Every Claude user right now:

@AjayReddyeh 👍👍

@KrishXCodes Hi @KrishXCodes , Plss check your dm....i hoe i get some suggestions from you...

Holy, this might genuinely change everything 😱 12M Context Window?? <5% Opus cost?? The numbers on this are unreal, 1000x less compute?? If this hold’s up it’s actually insane 🥶

@hrshtdotme @dhanushdotcodes Lfg!!! 🔥🔥

I got into a YC backed startup, btw.

@getnachonacho fr 🫡

@KrishXCodes Love this, curiosity-driven projects build real skills and stories that stand out more than just learning tools. That kind of initiative always pays off. Keep going, it compounds.

Best career decision I made wasn't learning a new framework. It was spending enough time building something I was curious about. Nobody asked me to build it. Nobody was waiting for it. But that project became the reason I got hired. Curiosity > strategy when you're starting out.