AI in a Nutshell - Week 32 - GPT-5, Grok Goes Wild, and Vogue Gets Roasted
This week, we had: GPT-5 go live, Grok go rogue, Vogue go full AI, some new AI tools, and a deepfake detector that might actually work.
Fellow human, it's week 32, and here's your no-fluff 4-minute version of what went down in the AI world in the past week.
If you're feeling lazy, here's the most important summary: OpenAI drops GPT-5 for everyone, Musk’s Grok ditches the morality filter, and Vogue swaps real models for AI (hint: people are not happy).
What You Must Know:
OpenAI Launches GPT-5 → The smartest ChatGPT yet, with four versions to fit your needs, now open to everyone. Some say it’s giving PhD-level answers without the tuition fees 😀
xAI Lets Users Make NSFW Content → Elon Musk’s Grok-Imagine tool goes rogue with no safety filters. Extrelemely unhinged. Expect art, chaos, and a moderation headache. Let me just say, you don't want your kids using this 🙏
Claude Opus 4.1 Drops → Anthropic’s latest Claude update crushes tough coding and reasoning tasks, keeping the AI race spicy, but some say it's a joke when compared side-by-side with GPT-5 🤡
Vogue’s AI Model Backlash → Vogue ran a campaign using AI-generated women instead of real models. Critics say it’s a blow to representation (and jobs) 😮
Deepfake Detector Hits 98% Accuracy → Scientists built a tool that can spot AI fakes in both video and audio. It could be huge for fighting misinformation, or we can wait a few weeks for a new AI tool that can bypass it 🥷
OpenAI Ends Google Search Result Experiment → ChatGPT shared convos were showing on Google. Then OpenAI gave it a quick “nope”, and took it off. Now what?
What’s Good to Know:
Meta AI Chats Might Now Always Private – Some Meta AI convos are public by default. Yes, even the personal ones where you used it as your therapist.
Kids’ Mood-Monitoring App Launches – Uses facial recognition and language analysis to alert parents. Helpful? Creepy? You decide.
New AI Tools Worth Knowing:
Grok Imagine → Create short, snappy AI-generated videos (hello, Vine nostalgia).
Gemini Storybook → Custom illustrated storybooks with audio narration.
Vibescan → Scans your AI-written codes for potential problems.
Droidrun → Grant AI native control of virtual & physical phones.
Testsprite → AI Agent for software testing.
Meme of the Week:
Kaizen Corner – What’s an AI Agent? And how’s it different from an AI Chatbot?
An AI agent is a system that combines decision-making with actions i.e it can do things for you.
Like → Reading emails, booking flights, or coding an app, all without you micromanaging every step like a deranged manager.
An AI agent is like a personal assistant who actually goes out, runs errands, and gets things done, while an AI chatbot is like a friendly neighbor who chats with you over the fence.
Question → What would you do?
If you were the CMO of Vogue, and you had two options:
Use human models and spend $6M on a campaign.
Use AI models and spend $6K on the same campaign.
Let's assume the campaign results would be the same. What would you do?
That’s your week in AI. Served short, sharp, and with just enough for you to not sleep off while reading it.
If you learned something, tell a friend. If you didn’t, blame yourself :)
Until next Sunday,
Kay, your fellow human.
P.S. If this email lands in spam/promotions/updates folder, that’s just your inbox trying to protect you from staying plugged in. Fix it.