Trump's AI Posting Spree

What we’ll cover today:

🤖 Trump flooded politics with AI images.

🧠 Anthropic made its model less annoying.

💼 A $1B startup wants to automate office work.

⚠️ An ex-OpenAI researcher says nobody is ready.

🤖 Trump Is Flooding Politics With AI

• Trump posted 43 AI-generated images in a single day, targeting critics, journalists and Iran.

• He also hinted at a 2028 run, even though a third term isn't legal.

• Several images placed him beside JFK and George Washington, turning AI into a political branding tool.

• AI is making it cheaper than ever to manufacture attention. Politics will only get noisier from here.

🧠 Anthropic Blinked First

• Anthropic launched Opus 5, a cheaper model that beats Fable 5 on several benchmarks.

• It also removed many of the restrictions developers complained about.

• Blocked prompts now fall back to a safer model instead of failing completely.

• AI labs are learning a simple lesson. The model that says "yes" more often usually wins.

💼 Coding might not be AI's Biggest Market

• Prentis is reportedly raising $100M at a $1B valuation to build AI office workers.

• Its agents are designed to handle paperwork, forms and repetitive business tasks.

• The startup claims it already has up to $50M in customer contracts, though its benchmark claims remain unverified.

• Everyone is building coding agents. The bigger business may be replacing office admin instead.

🛠️ TOOLS

🧠 BrainFeed – Turn articles into scrollable learning cards. →

🤖 Aymo – Compare 45+ AI models side by side. →

📱 Pushary – Approve AI agent actions from your phone. →

🎧 Heard – Hear smart summaries from coding agents. →

📚 RESOURCE

📚 He Risked $2M To Warn You About AI

Former OpenAI researcher Daniel Kokotajlo explains why he walked away from $2 million, why he believes superintelligence could arrive before the end of the decade, and why he thinks AI could automate almost every job. He also shares the one plan he believes could still keep AI safe.

Watch his warning →