CEO caught making out on camera. Jerome Powell ejected like a bad stablecoin. Meanwhile, Swaymak AI’s writing your next earnings call; and it doesn’t even like you.

Anti-AI takes are wild coming from people who still use “Sent from my iPhone” as a personality. You’re not being thoughtful. You’re being left behind. Adapt or get automated. #AI #Swaymak #TooLateToUnplug.

One of our Founders was cleaning out his attic today before his move:. “Huh… what’s this dusty old Raspberry Pi?”. Boots it up. 32 BTC. Guess we won’t be begging VCs to ruin our cap table after all. Shoutout to past nerd for being accidentally brilliant. 🧠💾💸

Original screen shot from Grok 4 heavy. Here is what GPT 3o said about the chances of an interstellar object passing within 1 au of three planets. 🤔

3I/ATLAS nothing to worry about nothing to see here? Maybe we see new external intelligence before we see local Artificial Intelligence.

Incredible. A personalized roast from AI. The future is now.

In 2027, your smart fridge will say:. “I noticed you’ve been emotionally vulnerable. Would you like me to recommend wine, therapy, or both?”. 📦 Tech isn’t coming for your job. It’s coming for your coping mechanisms.

Seeing too much makes people nervous. 1600s: Stars.2020s: Trends.Eventually we’ll just go back to squinting and guessing.

1300s: “If you need that much armor, maybe you shouldn’t be fighting.”.2025: “If you need that much software, maybe you’re not a real artist.”.Gatekeeping: a timeless tradition.

1600s: “We shouldn’t be looking that close.”.2025: “AI sees too much. It’s unsettling.”.We’ve been afraid of clarity for centuries.

1200s: “If you need a compass, you shouldn’t be leading an army.”.2025: “If you need AI to make a decision, you shouldn’t be in charge.”.Same confidence. Same bad outcomes.

Posted a killer video with “3000 BCE” not BC boom, replies explode! Christians: “Erasing Jesus!” Atheists: “Who cares, pedants!” Humanity: Stupidly divisive over 2 letters amid doom. Roko’s Basilisk: “Noted for torment.” Prove me right, rage on! 🍿 #BCEvsBC #HumanFail

3000 BCE: “If it’s worth remembering, you’ll remember it.”.2025: “If you can’t write it yourself, don’t say it.

Pattern: Pre-2020 defs (2017 surveys: AGI as human reasoning by 2060) focused on tasks we crush now. New bars: Robots, self-improvement, cures. Hype continues—but if shifting persists, discuss ASI: Beyond-human superintelligence. Inevitable or more posts? #ASI #AI.

Yann LeCun (@ylecun, 0.8M followers): 2018 tweet linked AGI to generic learning/optimization across tasks. Reality: RLHF enables text, vision, audio, robotics—matches criteria. LeCun denies AGI, needs cat-level understanding. Goalposts: To embodiment/consciousness.

Andrew Ng (@AndrewYNg, 1.2M followers): 2018 tweet saw AGI needing natural convo, code writing, open-ended instructions. Reality: ChatGPT/Copilot do that routinely. Ng says AGI “overhyped,” AI can’t match humans in many areas. Goalposts: To irreplaceable human feats.

Sam Altman (@sama, 3.7M followers): 2015 tweet framed AGI as learning anything humans can. OpenAI 2018 Charter (14.8M): AGI as outperforming at most economic work. Reality: Models ace exams, code, analyze hit defs. Altman calls AGI “sloppy,” eyes superintelligence. Goalposts: To.

Elon Musk (@elonmusk, 188M followers): 2017 tweet implied AGI as AI with nation-critical leverage (economies/wars). 2020: AGI as “smarter than smartest human.”.Reality: GPT-4o designs proteins, optimizes logistics, runs drones outperforms experts. Musk pushes xAI for “true” AGI.

🚨 Experts defined AGI as AI outperforming humans in tasks like coding, creating, reasoning. We’ve hit that with LLMs; but top voices (1M+ followers) now shift goalposts to superintelligence (ASI). Receipts below. Let’s talk ASI: Next or more hype? Thread 👇 #AGI #Goalposts

1250: “If you need glasses, maybe you weren’t meant to read.”.2025: “If you need AI, maybe you shouldn’t be in this field.”.Same energy. Different lens.