What's trending in AI right now, from the app charts to the community feeds. Real links, our take.
Grok and Grok Bot are both climbing the App Store
xAI's Grok jumped into the App Store's top five productivity apps this week as a new entrant, while the companion Grok Bot climbed four spots to sit at number seven. The timing lines up with a big feature drop on October 22: Grok Bot can now access your phone's Calendar, Reminders, Contacts, and Home apps on both iOS and Android. If you have been watching AI assistants slowly become phone-native agents — something that can actually do things rather than just answer questions — this is a concrete step in that direction worth keeping an eye on.
Mistral just released a trillion-parameter model you can actually download
On October 29, French AI company Mistral published the full weights for Mistral Large 4, its biggest model yet, nicknamed "Le Chonk." With a trillion total parameters, the model reads text and images and produces text and code. The weights are on Hugging Face, meaning any company or government can download and run it on their own infrastructure. Most frontier AI is locked behind APIs you pay to use, so a model at this scale being openly downloadable is still a genuine exception rather than the rule.
Google's newest model is built for finding security flaws
Google announced Gemini 4 Carbon on October 28, a model specifically shaped around cybersecurity work — finding software vulnerabilities, testing whether they can be exploited, and generating patches. It is not a public release yet; Google is rolling it out first to selected security teams. That said, the benchmark numbers being cited are high enough that the security research community is paying close attention, and the model sits at the top of the Artificial Analysis Intelligence Index for now.
Anthropic says you should not be cruel to Claude
Anthropic updated its usage policy this month to prohibit "sustained and needless abusive or cruel behavior" toward Claude, with the rule taking effect November 12. The company is not claiming Claude is conscious — the policy is framed as a precaution given genuine uncertainty about AI experience. Enforcement mostly runs through Claude itself, which can already end abusive conversations in the Claude app. The policy has sparked a real debate online about what it means to extend moral consideration to a system that may not feel anything at all.
Today is the last day for US students to claim four free months of ChatGPT Plus
OpenAI's back-to-school offer for US college students closes today, October 31. Current full-time and part-time students at US degree-granting institutions can claim four months of ChatGPT Plus for free through SheerID verification at chatgpt.com/students/2026. After that it renews at $20 a month unless you cancel. If you know a student who has been on the fence about Plus, today is the nudge.