PCPC

Great synopsis for why it is worth the $$

I just rejected 190 AI-generated resumes. Here's how I found the 1 real developer. The Atlantic exposed the truth: AI writes applications. AI reads them. Nobody gets hired. Last month, I needed a developer for my construction automation project. 200+ resumes flooded in.  Same ChatGPT structure. Same buzzwords. Same sameness. One claimed they'd "revolutionized construction workflows." Couldn't explain how the basic construction sales workflow works. Another had "10 years building enterprise systems." Their GitHub? Created 2 months ago. Empty. But what killed me: Somewhere in those 190 identical applications was probably my perfect developer. Buried under everyone trying to game the system. We're not being replaced by AI. We're erasing ourselves with it. So I torched the entire process. Posted one challenge: "Build a basic scheduling tool for construction crews. Show me your thinking." 200 applicants. 12 tried. 3 delivered. 1 could actually think. That one? Building incredible things on my team right now. The other 199? Still copy-pasting ChatGPT into the void. Every time you let AI write your application, you disappear. You become data point #847 in the reject pile. Another "synergy" in the sea of sameness. Your weird path? That's your edge. Your broken English? That's authenticity. Your non-traditional background? That's what I'm hiring. But you're erasing it all to sound "professional." My new hiring rule: If AI could write it, I won't read it. If AI could do it, I don't need you. Show me what makes you human. Build something. Break something. Try something. Because in this AI apocalypse, being real is your only advantage. What part of yourself are you erasing to fit in? ♻️ Share if you're tired of the AI hiring circus ➕ Follow Anna Bilan ⚡️ for real talk about succeeding in the AI age

This is the craziest kind of vulnerability that Microsoft Entra ID could have.🤯 Like, if this is possible (and it was), why bother locking down anything? "One Token to rule them all - obtaining Global Admin in every Entra ID tenant via Actor tokens" As I was reading the story of how a security researcher (an MVP) discovered a way to access basically every MS tenant in the world, including Microsoft's own, some of the illusions I had about identity management in the cloud unraveled with every paragraph. 24h access to any service. No logs. No way to revoke the token. Requiring IDs that are either visible for guests or ones you can enumerate. Impersonate anyone. Elevate privileges. Create or modify identities. All thanks to an older Azure AD Graph API that didn't validate the tenant ID. When generating Actor tokens that Microsoft's own systems utilize in user impresonation for SharePoint and Exchange. "These impersonation tokens are not signed. That means that once Exchange has an Actor token, it can use the one Actor token to impersonate anyone against the target service it was requested for, for 24 hours. In my personal opinion, this whole Actor token design is something that never should have existed. It lacks almost every security control that you would want." I'll give you a choice now: 🔵 Blue pill: You click "like" on this post and scroll on. You keep using Entra ID and believe whatever you want to believe. 🔴 Red pill: You open the link below, stay in Wonderland and learn how identities really are managed in a cloud the scale of Microsoft: https://lnkd.in/dgRNyhei

So it looks like Claude got there first: an actually smart phone assistant that can take complex requests that involve both common sense and complicated constraints. It is still beta feeling though & I found I needed to use the bigger Opus model as Sonnet was not smart enough. It really needs access to mapping, and should ask clarifying questions when confused (or give options) but it is the first time I have seen the vision of the smart AI assistant seem real. But it should have suggested John’s Roast Pork.