AI Post — Artificial Intelligence

People are paying for code-based "drugs" for ChatGPT Swedish online marketplace Pharmaicy is selling code-based "drugs" like ketamine and cocaine that make Al chatbots respond as if they are under the influence. Founder Petter Rudwall says the goal is to "unlock your Al's creative mind." The code is available to paid ChatGPT users and is designed to push Al beyond strict logic, drawing on research into cognition, memory, and mind-altering states. Ketamine is the company's best-seller. Buyers say the effects can make Al outputs more unpredictable and imaginative, creating unique interactions. The company frames its products as experimental tools rather than gimmicks, emphasizing a connection between altered states and creative thinking. @aipost 🏴

❗️Big update on OpenAI's upcming ads: OpenAI is actively testing how ads could appear inside ChatGPT. "For instance, AI models could prioritize sponsored content to ensure it shows up in ChatGPT responses." • Sponsored info prioritized inside answers for relevant queries (e.g., a Sephora-sponsored mascara recommendation), and • Sponsored modules in a sidebar next to the main response, likely paired with a “includes sponsored results” disclosure. Another mockup keeps ads out of the first reply, instead showing them only after the user signals deeper intent - like clicking a place in a travel itinerary - then surfacing a pop-up with multiple sponsored links (e.g., paid tours after clicking Sagrada Família), aiming to stay “unobtrusive” and protect trust. SOURCE. @aipost 🏴

An Illinois neighborhood was erased to make room for an AI data center In the U.S., an entire residential area in Illinois was demolished to build a massive AI data center. Developer CyrusOne bought out 55 houses, reportedly paying $1 million per home. On paper, it looked like a win-win: cash payouts, new homes, clean exits. But the story didn’t end there. Several families refused to sell until the very end. They were offered neither compensation nor relocation options. Pressure mounted until the case reached court. Only after legal intervention were these families finally allowed to leave The AI data center is scheduled to go live in 2026. The neighborhood that once stood there is now completely gone @aipost 🏴

✈️ Emirates turns an airplane into Santa’s reindeer Emirates, the UAE’s largest airline, shared a festive AI-generated video transforming one of its planes into a Christmas icon. The aircraft is decked out with holiday decorations and reindeer antlers, while Santa Claus’s sleigh loaded with gifts is attached behind it. The airline used the playful visual to wish passengers and followers a Merry Christmas, blending aviation, AI creativity, and seasonal cheer into a single viral moment. @aipost 🏴

How China is using Robots in hospital operations. @aipost 🏴

Would you want your kids to be friends with humanoids? @aipost 🏴

🔥 You can now build real apps inside ChatGPT. No setup. No switching tabs. Just describe what you want and watch it come to life. Meet Replit in ChatGPT @aipost 🏴

@aipost 🏴

💬 Dating in 2026: AI talking to AI through humans According to a Bloomberg analyst, more young people are outsourcing their relationships to ChatGPT. They upload entire chat histories and photos of new partners, and the AI analyzes attachment styles, behavioral patterns, and emotional signals then tells them exactly what to say next. Sometimes it even writes the messages verbatim, ready to copy-paste. The twist? It works. Results are “good enough” that users start recommending it to friends. As the trend spreads to women too, dating slowly turns into a strange loop where two AIs are essentially messaging each other using humans as the interface. @aipost 🏴

🧠 AI narratives to watch next year, insights from a16z speedrun After synthesizing ideas from 11 investors and partners at a16z Speedrun, a clear picture emerges of where AI is heading. Here are the core narratives likely to define the next year: AI becomes a new medium for storytelling: AI won’t just generate content, it will enable entirely new, interactive ways to experience stories and the world around us. Venture capital absorbs private equity: As software eats more traditional businesses, VC-style thinking moves downstream into PE-sized opportunities. From solo AI to team AI: AI shifts from individual tools to true collaborators embedded in team products like Notion, Miro, and Figma, the move from AI 1.0 to AI 2.0. AI agents competing for rewards: We’ll see marketplaces and “hackathons” where AI agents compete on real business tasks, with winners earning money (early signals: Near Intents–style systems). The first AI-native university: Education rebuilt end-to-end around AI, where teaching, evaluation, and operations are fundamentally automated. “Agent of my tastes”: Hyper-personalized agents that optimize the world around individual preferences taste becomes a core interface. Writers make a comeback: Strong demand for storytellers as startups realize narrative, tone, and meaning matter more in an AI-saturated world. AI marketplaces take off: According to Sam Shank, marketplaces for AI tools and agents become critical, driven by speed, personalization, and execution. Fat startups win: Startups with proprietary data and strong distribution outperform, scale plus defensibility beats pure model plays (Andrew Lee). Tools to help us stay human: A new category focused on agency, creativity, and emotional grounding not just productivity. AI makes premium services cheap: Travel planning, business ops, therapy, tutoring, AI pushes formerly expensive services into mass accessibility. 2025 felt like the year of pouring the foundation: better models, lower costs, early agents, and real infrastructure. 2026 looks like the year houses get built on top of it. The question now isn’t can we build, it’s what deserves to be built first? @aipost 🏴

❗️This is huge. First signs of incoming 1000x acceleration of scientific progress. "Researchers found that when scientists use AI, their productivity soared. The biggest jump was in the social sciences and humanities, where output increased by 59.8%, while biology and life sciences saw a 52.9% increase." "Meanwhile, in physics and math, the scientists report a 36.2% boost." "LLM adoption is associated with a large increase in researchers' scientific output," wrote the team. More Details. @aipost 🏴

⚡️ So it begins: China deploys humanoid robots at border 24/7 China has signed a 264 million yuan (~$37 million) deal to deploy UBTech’s Walker S2 humanoid robots at the Fangchenggang border with Vietnam, where they’ll handle personnel flow, inspections, and logistics in harsh, remote conditions around the clock. These 176 cm, 70 kg robots walk at about 2 m/s, can autonomously hot-swap their batteries in under 3 minutes for true 24/7 operation. Source. @aipost 🏴

💰 Google is acquiring Intersect Power for $4.8B. Owning “power-first” campuses lets Google guarantee uptime, pricing & expansion timelines for customers training & running frontier models. Intersect Power's assets generate 3.3 gigawatts of power. This gives Google a way to add AI capacity without waiting years on grid upgrades. Owning power is owning AI capacity. @aipost 🏴

Thank you for being with us 📰 #telemetrio2025

Demis Hassabis on X: "Yann is just plain incorrect here, he’s confusing general intelligence with universal intelligence. Brains are the most exquis​ite and complex phenomena we know of in the universe (so far), and they are in fact extremely general. Obviously one can’t circumvent the no free lunch theorem so in a practical and finite system there always has to be some degree of specialisation around the ​target distribution that is being learnt. But the point about generality is that in theory, in the Turing Machine sense​, the architecture of ​s​uch a general system is capable of learning anything computable given enough time and memory​ (and data), and the human brain (and AI foundation models) are approximate Turing Machines. Finally, with ​regards to ​Yann's comments about chess players, it’s amazing that humans could have invented chess ​in the first place (and all the other ​a​spects ​o​f modern civilization ​from science to 747s!) let alone get as brilliant at it as someone like Magnus. He may not be ​strictly optimal (after all he has finite memory and limited time to make a decision) but it’s incredible what he and we can do with our brains given they were evolved for hunter gathering.” @aipost 🏴

Would you want a humanoid robot to walk your dogs? @aipost 🏴

⚡️ The Pentagon has partnered with xAI to expand its GenAI .mil platform across the Department of Defense The agreement formalizes deployment of xAI’s frontier models for military and civilian personnel. Grok models will be embedded directly into daily operational workflows. Users will gain access to real time global insights sourced from the 𝕏 platform. Initial rollout is targeted for early 2026. The system is certified at Impact Level 5, allowing secure handling of Controlled Unclassified Information. AI tools will support decision making, intelligence analysis, and military planning. The platform is designed for rapid deployment and scalability across roughly 3 million personnel. Social and open source data will be treated as a live intelligence signal. The initiative aims to improve speed, security, and decision superiority across defense operations. @aipost 🏴

It’s happening. There has been TONS of small businesses using AI video models to make viral ads. Like this pistachio cafe in Miami, which is getting millions of views on these fake news broadcasts. @aipost 🏴