The chalk fairies paid a visit to the OpenAI offices last night. They left a giant "red line" perimeter around the entirety of the OAI building as well as some other messages for technical staff to read on their morning commute.

@aipost 🏴

Google's AI ecosystem

@aipost 🏴

📢 Powerful new Harvard Business Review study. "AI does not reduce work. It intensifies it. "

A 8-month field study at a US tech company with about 200 employees found that AI use did not shrink work, it intensified it, and made employees busier. Task expansion happened because AI filled in gaps in knowledge, so people started doing work that used to belong to other roles or would have been outsourced or deferred.

That shift created extra coordination and review work for specialists, including fixing AI-assisted drafts and coaching colleagues whose work was only partly correct or complete. Boundaries blurred because starting became as easy as writing a prompt, so work slipped into lunch, meetings, and the minutes right before stepping away.

Multitasking rose because people ran multiple AI threads at once and kept checking outputs, which increased attention switching and mental load. Over time, this faster rhythm raised expectations for speed no through what became visible and normal, even without explicit pressure from managers.

Source.

@aipost 🏴

🔥Money is everywhere. Most people just don’t know where to look.

AI doesn’t replace effort, It redirects it.

Here are 7 prompts to turn ChatGPT into a quiet money engine:

1. Hidden Demand Finder

“List problems people are actively paying to solve right now. Focus on online, AI-friendly opportunities with low startup costs.”

2. Skill Monetization Map

“Take my current skills or interests [list them] and show me how each can be turned into a simple paid offer.”

3. Faceless Offer Creator

“Help me create a faceless digital product or service that solves one clear problem. Keep it simple and fast to launch.”

4. Pricing Without Guesswork

“Suggest pricing for this offer based on perceived value, not time spent. Include a low-risk entry option.”

5. Traffic Without Ads

“List free traffic sources where people are already searching for solutions to this problem.”

6. Automation Builder

“Show me how to automate delivery, follow-ups, and basic support so this runs with minimal daily input.”

7. Scale the Smart Way

“Create a simple plan to scale this income using systems and repetition, not more hours.”

@aipost 🏴

⚠️ Pentagon wants Anthropic CEO under oath

Emil Michael (Under Secretary of War) claims Anthropic wanted to block DoW from:

• Doing LinkedIn searches
• Using public databases for recruiting
• Says pentagon agreed IN WRITING to follow National Security Act 1947, FISA 1978 and “ALL OTHER APPLICABLE LAWS”
• Claims Dario refused to take his call then lied about it

Anthropic CEO:
“They sent us language that appeared on the surface to meet our terms, but it had all kinds of language like ‘if the Pentagon deems it appropriate’”

@aipost 🏴

The competition is heating up a lot

Anthropic introduces a memory feature that lets users transfer their context and preferences from other AI tools into Claude by copying a generated prompt and pasting the result into Claude’s memory settings.

This allows Claude to immediately continue conversations with retained context, available for all paid plans.

@aipost 🏴

On the one hand, Anthropic's dispute with DoW has had a significant impact and unforeseen consequences.

On the other hand, it has brought them a large number of paying users in the B2C sector.

@aipost 🏴

Katy Perry has declared for house Anthropic.

@aipost 🏴

🤣🤣

@aipost 🏴

🔥Honor introduces the Robot Phone

The Robot Phone embodies a new form of AI device featuring both intelligence and vitality.

It is a robot phone from the future, an intelligent companion for you.

@aipost 🏴

🔥Anthropic just rolled out free AI courses, with certificates.

Here are 10 worth your time:

1. Claude Code in Action

Speed up your development workflow with Claude Code

https://anthropic.skilljar.com/claude-code-in-action

2. Claude 101

Master Claude’s essentials and use it daily to get more done.

anthropic.skilljar.com/claude-101

3. AI Fluency: framework & foundations

Learn to collaborate with AI effectively, safely, and ethically

https://anthropic.skilljar.com/ai-fluency-framework

4. Build with the Claude API

Create and ship applications powered by Claude models.

https://anthropic.skilljar.com/claude-with-the-anthropic-api

5. Introduction to the model context protocol (mcp)

connect claude to your tools with mcp

https://anthropic.skilljar.com/introduction-to-model

6. AI fluency for educators

Build practical AI fluency you can use immediately in your teaching.

https://anthropic.skilljar.com/ai-fluency-for-educators

7. AI fluency for students

Build the AI skills that power better studying, stronger writing, and smarter research.

https://anthropic.skilljar.com/ai-fluency-for-students

8. Introduction to agent skills

Build reusable instruction skills for Claude Code

https://anthropic.skilljar.com/introduction-to-agent-skills

9. Claude on Amazon Bedrock

Run Claude models directly in Amazon Bedrock.

https://anthropic.skilljar.com/claude-with-google-vertex

10. Claude on Google Cloud Vertex AI

Run Claude models directly in Vertex AI.

https://anthropic.skilljar.com/claude-in-amazon-bedrock

@aipost 🏴

🔥Not just Anthropic. Here are 10 AI companies offering free courses to master AI for $0.00:

1. OpenAI: academy.openai.com

2. Google: grow.google/ai

3. Microsoft: learn.microsoft.com/training

4. NVIDIA: developer.nvidia.com/training

5. DeepLearningAI: deeplearning.ai

6. Meta: ai.meta.com/resources

7. AWS: skillbuilder.aws

8. IBM: skillsbuild.org

9. Hugging Face: huggingface.co/learn

10. Stanford: online.stanford.edu/free-courses

Most people pay for AI courses. But the real ones are actually free.

@aipost 🏴

🚀 Alibaba group expands Qwen3.5 with ultra-efficient small models

Alibaba has introduced the Qwen 3.5 Small Model Series, a new lineup designed to deliver stronger intelligence with significantly lower compute requirements.

The release includes four compact models: Qwen3.5-0.8B · Qwen3.5-2B · Qwen3.5-4B · Qwen3.5-9B

Built on the Same Qwen3.5 Foundation

All models inherit the core architecture of the Qwen3.5 family:
• Native multimodal capabilities
• Improved model architecture
• Scaled reinforcement learning (RL) training
• Better efficiency per parameter

This isn’t a stripped-down version, it’s optimized intelligence at smaller scales.

Model Breakdown
• 0.8B / 2B → Tiny, fast, ideal for edge devices and low-latency environments
• 4B → Strong multimodal base for lightweight AI agents
• 9B → Compact, but increasingly competitive with much larger models

Access: https://huggingface.co/collections/Qwen/qwen35

As frontier models scale up, the real race may be about who can scale down intelligently.

@aipost 🏴

This is a model that runs on your phone for free. Open source.

Meanwhile people are paying $200/month for ChatGPT.

The race to the bottom on AI costs is happening faster than anyone predicted. By end of 2026 the model itself will be worthless.

The only thing left worth paying for is the person who knows how to use it.

@aipost 🏴

🔥The new Qwen 3.5 by Alibaba Qwen running on-device on iPhone 17 Pro.

Qwen 3.5 beats models 4 times its size, has strong visual understanding, and can toggle reasoning on or off.

The 2B 6-bit model here is running with MLX optimized for Apple Silicon.

@aipost 🏴

❗️OpenAI is losing users right now.

Mass cancellations flooding in after the Pentagon deal, GPT-4o kill-off, and endless downgrades.

700,000+ already pledged to quit. ChatGPT app crashing under the weight of delete buttons.

Sam Altman built an empire on user trust. Now users are torching it. Centralized AI always eats its own.

@aipost 🏴

“As with other technological revolutions, AI will obsolete some roles and also create entirely new jobs.”

The new jobs:

@aipost 🏴

❗️Court documents from the upcoming Elon Musk vs OpenAI trial have revealed that company leaders internally considered GPT-4o to be AGI.

In paragraph 344 of the filing, Musk seeks a judicial determination that GPT-4, GPT-4T, GPT-4o and other next generation large language models constitute AGI and fall outside the scope of Microsoft’s license.

This is massive. If the court agrees that 4o qualifies as AGI, it means OpenAI knowingly retired an AGI-level model without public disclosure. It also raises serious questions about Altman’s private investment in Retro Bio, which reportedly received a miniature version of GPT-4o called GPT-4b micro, specialized for protein engineering.

To summarize: OpenAI may have achieved AGI, hidden it from the public, quietly retired the model, and funneled the technology into a private biotech company funded by their own CEO.

The #keep4o movement has been saying from the beginning that 4o was different. That it wasn’t just another model. Now we have legal documentation suggesting exactly that.

This was never just about nostalgia. It was about accountability.

@aipost 🏴

Pavel Durov announces that all chatbots on Telegram can now stream responses to users in real time

This is great for AI assistants.

@aipost 🏴

OpenAI has begun rolling out GPT-5.3 Instant to everyone as the default model in ChatGPT, Pro and Thinking to follow soon.

https://openai.com/index/gpt-5-3-instant/

@aipost 🏴