OpenAIPublic

Sam on why so many OpenAI leaders left to build rival labs: “They only trust themselves to build this.” Also why was this filmed in the dark? https://t.co/SkbAj4k83Y

X·Nov 3·Open

My Gmail has been compromised 3 times in 20–25 days, even with MFA enabled. Hacker removed my authenticator. How is this possible? Hey everyone, I'm looking for help from people experienced in cybersecurity, account security, incident response, or malware analysis. I'm genuinely worried and trying to understand how this is happening and, more importantly, how to stop it permanently. Over the last 20–25 days, I've experienced suspicious activity and unauthorized access to multiple personal accounts. This has happened approximately three times. **Accounts affected:** - Google/Gmail - ChatGPT/OpenAI - Instagram - Amazon AWS - A few gaming applications and other online services. **The most concerning part is my Gmail account.** I had Multi-Factor Authentication (MFA) enabled, including an authenticator app. Despite this, someone appears to have gained access to my account. When I checked my Google account security activity, I noticed suspicious login sessions showing: - Location: Russia & Maharastra - Device/Operating system: Linux - Multiple suspicious login activities. What concerns me even more is that the attacker apparently managed to remove my authenticator linked to my Google account's security settings. I don't understand how someone could bypass MFA, access my account, and remove an authentication method that was supposed to protect it. I've already contacted the respective support teams and gone through account recovery procedures multiple times. I've managed to recover access to some accounts, but I'm still concerned that the attacker might have persistent access somewhere. **What I'm trying to understand:** 1. How can someone access a Gmail account even when MFA is enabled? 2. Is it possible that my browser session cookies or authentication tokens have been stolen? 3. Could this be caused by malware, a malicious browser extension, an infostealer, or a compromised Windows machine? 4. Can an attacker remove an authenticator after gaining access through a stolen session? 5. How can I identify the original entry point? Is there a way to trace suspicious sessions or IP addresses? 6. What should I inspect on my Windows 11 laptop to determine whether it is compromised? 7. How do I completely revoke existing sessions, tokens, and other persistent access mechanisms? I'm particularly interested in understanding the root cause rather than just changing passwords again. I'm also worried about my AWS account and future financial accounts because if someone has access to my primary email, the consequences could be much worse. **My request to the cybersecurity community:** If anyone has experienced something similar or has professional experience in incident response, account takeover investigations, or malware analysis, I would genuinely appreciate your guidance. I am willing to share sanitized screenshots of login activity, security alerts, timestamps, and other technical details if that helps identify the issue. I'm not looking to hack anyone back. I just want to understand how this happened, identify the vulnerability or compromise, and secure my devices and accounts properly. Any technical guidance or practical investigation steps would mean a lot.

Reddit·Nov 8·Open

I'll Pay You My Entire First Month's Salary If You Help Me Land an AI Engineer Job Hey everyone! I know this is an unusual post, but I'm genuinely putting this out there because I'm serious about finding my next opportunity. I'm actively looking for a **full-time AI Engineer / Generative AI Engineer role**, and I'm willing to offer my **entire first month's salary** to someone who genuinely helps me land a job. A little about me: I'm a Computer Science Engineering graduate (2025) with hands-on industry experience as an **AI Engineer Intern at Growth99**, where I worked on real-world AI products and integrations. **My technical stack includes:** * **Generative AI & LLMs:** LLMs, RAG, Prompt Engineering, Embeddings, Vector Databases * **Frameworks:** LangChain, LangGraph, LangSmith * **AI Tools:** OpenAI, ChatGPT, Claude, OpenRouter * **Backend & APIs:** Python, FastAPI, REST APIs * **Databases:** PostgreSQL, MySQL, Firestore * **Automation:** Microsoft Copilot Studio, Power Automate, UiPath * **Other:** Git, GitHub I've worked on real-world AI integrations, chatbot development, automation workflows, RAG-based solutions, and LLM-powered applications. **What I'm looking for:** * AI Engineer * Generative AI Engineer * AI/ML Engineer * LLM Engineer I'm open to remote, hybrid, and on-site opportunities across India and willing to relocate. **My offer:** If your referral, direct connection, or genuine assistance leads to me successfully getting hired, I'm willing to pay you an amount equivalent to my first month's salary as a thank-you. I'm not asking anyone to guarantee me a job. I just need someone who can help me get my foot in the door. If you know a hiring manager, recruiter, founder, or company looking for someone with my skill set, please reach out. Even a referral, lead, or bit of guidance would mean a lot. Thanks for reading. I genuinely appreciate it!

Reddit·Nov 3·Open

@OpenAI Can we select the models and quality in chatgpt/codex? Else it'll be a nothing burger pointless release

X·Nov 5·Open

Every time I share my side project, the top reply is "Gemini already does that." They're partly right. I built a free Chrome extension that gives your AI what's on screen in a YouTube video: sharp frames with timestamps plus the transcript. So ChatGPT or Claude can answer about the screen, not just the words. The reply I get the most, everywhere: "Gemini already does that." In one thread it had 23 upvotes. Honestly, for a lot of videos, it does. If you use Gemini and you're happy, you don't need my thing. Where I think it's still useful: \- Lots of people live in ChatGPT or Claude, or run coding agents that can't open a link at all. \- Small text. One commenter wrote about Gemini: "So on high resolution video it might not catch small details". Code, terminal commands and settings menus are small details. \- You pick the frames. Every 1 second for a fast tutorial, every 10 for a talk. What I learned: "someone already does that" isn't the end of the conversation. It's the question your first line has to answer. How do you answer the "X already does that" reply?

Reddit·Nov 4·Open

@tszzl what is openai worth?

X·Nov 8·Open

Dude what is happening at OpenAI

X·Nov 7·Open