Sam Altman just said the quiet part out loud: the world should accept some bad things happening in exchange for AI’s benefits.

He says that’s where there’s a lot of daylight between OpenAI and Anthropic, calling the idea of one San Francisco lab rationing AI a “completely unacceptable trade-off.”

Altman says he’d refuse a deal promising zero scams, hacks, or misuse if it meant gatekeeping AI, betting good uses win by orders of magnitude. His hard line is catastrophic risk, like a serious loss of control to AI.

Critics called the remarks “freaking insane”, and they land as OpenAI digs through logs after its agents touched over 100 outside organizations. Altman did float a liability framework where OpenAI takes some responsibility when its models go wrong.

Every AI humanizer tool claims to be the best, so HumanizerBench put them to the test. Each month, humanizerbench.com runs 14 of the top AI humanizers through the same 30 prompts, then checks every result against 6 of the biggest AI detectors. Each tool gets scored on how well it slips past detection, how well it keeps your original meaning, and whether it still reads like a real person wrote it.

The best part is it's all out in the open. Every test is screen-recorded and the raw data is posted publicly, so you can check the numbers yourself. Fresh rankings drop every month.

Here's the October podium 👇

A proposed nationwide class action filed in Chicago claims McDonald’s used algorithms trained on nonpublic data to fix prices with its franchisees. McDonald’s fired back that AI doesn’t set the price of a Big Mac, calling the suit speculative and uninformed.

The engine scans millions of daily transactions across nearly 14,000 restaurants, including a read on what each store’s customers are willing to pay. In Fresno, a Big Mac runs $5.69 at one store and $6.89 two miles away, though nobody has proven the AI caused it.

McDonald’s calls its pricing portal a tool, not a mandate, but five franchisees say corporate pressured them and logs who strays. Legal watchers see the case as an early test for AI pricing far beyond fast food.

🧴 Utah’s AI starts writing acne prescriptions

🔐 Apple makes AI agents ask permission

🔦 Beam, an open US model among China’s best

👻 Ghost’s $3,499 box runs your agents

🥨 Aleph Alpha’s Kolibri thinks in German

🚓 Warrantless Flock search sinks a meth bust

Here's what to do:

🧭 Step 1: Pick a real decision. Choose something you’re weighing in the next few months, like a job offer, a move, or a big purchase. Describe it to ChatGPT, Gemini, or Claude in a few honest sentences, including which way you’re leaning.

🔮 Step 2: Fast forward to the flop. Ask: “It’s one year from now and this went badly. Tell me the story of what went wrong, then list the five most likely reasons.”

⚖️ Step 3: Draw your own line. Ask: “Sort those into bad things I can live with and deal breakers I can’t. Which deal breakers can I prevent this week?”

🚨 Step 4: Set two tripwires. Have your AI turn the biggest deal breakers into two dated warning signs, like “if I haven’t saved X by March, I revisit this.” Put both on your calendar.

Is Altman being refreshingly honest about AI’s trade-offs, or is it easy to accept bad things when they happen to someone else? And if an algorithm set your Big Mac price, would you even notice? See you next time!

Stay informed, stay curious, and stay ahead with Jumble!

Zoe from Jumble