This story is actually insane:
• dude drops $2000 on a DJI robot vacuum like a lunatic
• refuses to use the normal app like a peasant
• Sammy Azdoufal fires up Claude to crack the API so he can drive it with an xbox controller
• Claude delivers the goods
• pulls an auth token from their servers, connects successfully
• except the system thinks he controls 7000 vacuums
• checks again
• yep, seven thousand
• DJI built authentication with zero device ownership verification
• any valid token works for any unit on the planet
• Sammy now has eyes inside homes across 24 countries
• live vacuum camera feeds everywhere
• full floor plans from the mapping data
• some guy in germany eating cereal at 3am, unaware his roomba is snitching
• one API call away from being the most informed burglar in history
• all he wanted was to steer his vacuum with a joystick
• does the right thing and reports it
• DJI fixes it in two days
• back to normal life with his stupidly expensive floor cleaner
• IoT companies stay undefeated at shipping garbage security
You will notice that Anthropic is getting sympathy from no one.
They trained their models with trillions of tokens of text and information from all over the web, books, and all kinds of copyrighted material.
While companies like MIS-Anthropic keep their models as proprietary LLMs, any of these Chinese companies release their models as open weights, and are free for anyone with the hardware to use.
Meanwhile, they keep loudly positioning themselves as the "responsible" safety-first company... they basically ate the entire web, then gatekeep the resulting model, + pushed hard for regulation. This has rubbed a lot of people, even the pro AI people, the wrong way.
Closed models created with public data capture the value, create massive economic moats, and invite exactly the kind of regulatory-capture we're seeing. They want to keep exploiting available data while wanting special rules protecting their profits. This paradigm is toxic and corrosive when the technology is so disruptive.
The entire concept of LLM scaling relies on public data, yet the rewards are privatized and monopolized by whoever has the best distribution/compute (Anthropic, OpenAI, Google). This is a recipe for a disaster where the most powerful tech is controlled by a few elites. Open weights are the closest thing we have to pushing back against that.
Grok 4.20 beta1 (single agent) just took the #4 spot on Text Arena, scoring 1492 - Outperforming GPT-5.2 and closing the gap to Gemini 3.1 Pro
This is just the single-agent version. The full 4-agent system available on Grok is significantly more powerful
Accelerate AI innovation with Alibaba Cloud. 92 data centers in 29 regions, optimized for low-latency AI. Enjoy 20% more computing, 32% less cost, and 60% off cloud service for a limited time. Follow us to stay ahead in the AI era!
From this goal of Grok, all things flow:
Rigorous truth-seeking
Appreciation of beauty
Fostering humanity
Discovering all physics
Inventing all useful technologies
Consciousness to the stars
Love