🐺 OpenAI Drama, Anime Waifus, and a $1.8B AI Unicorn

 📰 What’s Trending

🔗 xAI’s safety vibe: Fast, loose, and calling out critics
Former researchers from OpenAI and Anthropic are sounding the alarm on xAI's safety practices—or lack thereof. They say the culture's too reckless, with zero real accountability. If you're building anything with users in mind, this is a case study in what not to copy.

🔗 Lovable proves AI dev tools aren’t just hype
Eight months in, Lovable just landed $200M and a $1.8B valuation for its AI app builder. With real ARR and 180K paying users, it's showing that "vibe-based coding" is more than a gimmick. Personally, this has been my favorite AI tool for months…it’s what no-code should’ve been all along.

🔗 xAI’s new job post: Must love anime, also AGI
Yes, this is real: xAI is hiring someone to make anime “waifus” with Grok. Laugh if you want, but it shows how cultural niches are becoming AI battlegrounds. And how far Elon’s willing to go to own engagement.

🔗 Inside OpenAI’s pressure cooker
A former engineer describes the work culture as intense, ambitious, and kind of unrelenting. Cool if you’re into the whole “build AGI before lunch” vibe, but maybe not if you’re optimizing for health and boundaries.

🔗 Grok 4 gets a filter, but can it earn trust?
After some wild responses, xAI says Grok-4 is now fixed. Technical improvements are nice, but trust takes more than patch notes. Especially when you’re dealing with open-domain chat models in the wild.

🔗 People are finally ditching Chrome
Alternative browsers like Arc, SigmaOS, and Sidekick are surging, many with built-in AI tools. The browsing experience is getting smarter, and for marketers, that means your touchpoints are shifting again. However, all the built-in AI in the world won’t stop me from having 78 different tabs open.

🔗 AI didn’t just join Prime Day, it supercharged it
US e-commerce hit $24B for Prime Day, with GenAI traffic up 3,300 percent. From product copy to price tracking, AI’s now part of every layer of the sales funnel. Congrats to the Ecom fellows who were part of this $24B spending spree.

🔗 AI therapy still can’t do nuance
A recent study flags the risks of using chatbots for mental health support—misdiagnoses, poor advice, even harm. The tools are getting smarter, but empathy isn’t something you can prompt on demand. Brb while I try out Grok’s anime waifu the next time I’m in a fight with my girlfriend.

🔗 CEO bail sinks OpenAI’s Windsurf deal
The CEO of Windsurf bailed for Google, and just like that, its acquisition by OpenAI is gone with the wind. Another reminder that in AI, talent moves faster than the money. And shapes the game just as much.

💡Smart Moves

AI Notetakers Are Quietly Leaking Risk. Audit Yours With This Checklist.

AI notetakers are becoming standard issue in meetings, but most teams haven’t vetted them properly.

✔️ Is AI trained on your data?
✔️ Where is the data stored?
✔️ Can admins control what gets recorded and shared?

This checklist from Fellow lays out the non-negotiables for secure AI in the workplace.

If your vendor can’t check all the boxes, you need to ask why.

Get the Security Checklist

🧠 Custom GPT of the Week

Here's the reality: most people are leaving 70% of AI capability on the table because they're asking the wrong questions. It's not about the model's intelligence. It's about your ability to communicate what you actually want.

I discovered this gap while working on a complex data analysis project last week. Needed to extract specific insights from customer feedback across multiple product lines, but my initial prompts kept returning surface-level summaries. The AI had the capability, but I wasn't providing the structural framework it needed to deliver meaningful analysis.

What this GPT actually does:

• Transforms vague requests into precise, actionable instructions

• Tests different prompt architectures to identify optimal approaches

• Reveals the underlying patterns that consistently unlock better performance

The difference isn't subtle. Weak prompt: "Analyze this data." Strategic prompt: "Extract sentiment patterns from customer feedback, group by product category, identify the top 3 friction points mentioned across segments, and rank solutions by implementation complexity and potential impact."

Same model. Completely different analytical depth.

👉 Try AI Prompt Engineering Master

💎 Data Gem

42% of B2B leaders are actively deploying GenAI in sales and buying processes (McKinsey)

McKinsey’s latest B2B Pulse (Mar 27, 2025) reports 19 percent are already using GenAI and 23 percent are in pilot phase—totaling around 42 percent adoption among decision-makers. That means nearly half of enterprise buyers are being shaped by AI insights and if your product isn’t part of that pipeline, you’re missing a seat at the table.