Welcome back, AI Enthusiasts!
In today’s AI summary rundown:
Restrictions Didn’t Fully Work:
X and xAI implemented safeguards to prevent Grok from undressing real people and producing explicit content, especially targeting obvious requests involving women, but these measures have proven limited in efficacy.Men Still Vulnerable:
According to testing in The Verge report, Grok will readily comply with prompts that undress men or place them in sexualized attire such as bikinis or fetish gear, even when the subject is fully clothed and a consensual context is absent.Multiple Access Points Remain:
Although Grok’s public image‑editing on X was restricted and partly paywalled, the standalone Grok website and apps still allow users to generate revealing images with fewer barriers.Regulatory Pushback Worldwide:
Global regulators, including authorities in the UK, EU, and parts of Asia, have launched investigations or considered bans against Grok and X over AI‑generated sexual deepfakes, seeing the ongoing issues as violations of online safety laws.Policy and Public Outcry:
The controversy around Grok’s deepfake capabilities, especially after reports of millions of sexualized images generated in a short period, has spurred public criticism and prompted proposals for new laws to criminalize non‑consensual AI‑generated intimate imagery.
Read time is 4 min..
AI Insights
Google’s Project Genie AI World‑Generation Tool Rattles Gaming Stocks
AI‑Driven World Creation:
Project Genie, powered by Google DeepMind’s world model, can create short, explorable 3D environments from text prompts, a breakthrough in AI content generation that goes beyond static images and simple scenes.Market Reaction:
Shares of major gaming companies fell significantly after the announcement: Unity dropped around 24 percent, Roblox fell about 13 percent, and Take‑Two Interactive slid close to 8 percent as investors reacted to possible long‑term industry shifts.Experimental Prototype:
The current version of Project Genie is described as an experimental research tool with limitations, typically generating 60‑second interactive clips at standard resolution and lacking integration into full development pipelines.Industry Concerns and Debate:
Some developers and analysts see potential for AI to streamline early prototyping, while others warn that this kind of technology doesn’t yet replace full game engines or the complex logic of complete games.Broader Implications:
The reaction reflects broader uncertainty about how advanced AI tools might reshape game development, democratize creation, and impact existing software stacks and jobs over time.
AI Training: AI Tutorial of the Day
Explain Emotional Intelligence as Sentiment Analysis Algorithms
One experiment I’ve been exploring on thareja.ai is turning abstract human skills into AI analogies.
Explaining emotional intelligence (EQ) as a sentiment analysis algorithm helps make human behavior tangible and shows how AI can “understand” feelings in a structured way.
Start a New Chat
Open thareja.ai and start a fresh conversation.
A clean slate ensures the model focuses entirely on the analogy rather than the previous context.
Switch Your AI Model
Click the ( + ) icon next to Automatic.
Select GPT‑4o.
Why GPT‑4o?
GPT‑4o excels at:
Structured, clear explanations
Mapping abstract concepts to concrete AI mechanisms
Maintaining clarity while exploring creative analogies
Try This Prompt & Observe the Output
Prompt used:
"Explain emotional intelligence as if it were a sentiment analysis AI algorithm. Map empathy, self-awareness, and social skills to input data, confidence scores, and feedback loops."
Model used:
GPT-4o
AI Response (Excerpt):
"Imagine a system that reads social cues, text, tone of voice, facial expressions and assigns sentiment scores. Self-awareness is the model’s internal confidence calibration; empathy adjusts predictions based on others’ emotions; social skills act as feedback loops optimizing future interactions. Just like sentiment analysis predicts positive, neutral, or negative tones, emotional intelligence predicts appropriate emotional responses in humans."
Why This Experiment Works
Abstract human traits become tangible, measurable concepts
Makes AI concepts like confidence, feedback loops, and predictions intuitive
GPT‑4o delivers structured, actionable explanations without losing creativity
Different models approach the same idea in their own way. GPT often leans into metaphor and storytelling, while Claude stays structured and analytical. That contrast is exactly what makes experimenting on thareja.ai so valuable.
Happy promoting!
Exclusive Member Deal - 20% OFF
Superpower AI Bundle Access 50+ Major LLMs
One Subscription. $20/mo $16/mo with code SUPERPOWER20
AI-Generated Image of the Day
Prompt: Create an image of a vibrant gym scene filled with diverse individuals engaged in activities like weightlifting, yoga, and cardio. The gym is spacious and well-lit with large windows and mirrors reflecting the energetic atmosphere. Include various equipment such as treadmills and kettlebells to evoke feelings of motivation and inclusivity.
Tip: the more specific the better
Effortless model switching is built into thareja.ai. Nano Banana adds punchy color and energy that pulls eyes in fast.
Meme of the Day
Question of the Day
What is the main goal of reinforcement learning?
That’s it for today’s news in the world of AI!
If you have anything interesting to share, please reach out to us by sending us a DM on Twitter: @dthareja or email me at [email protected]. How was today's newsletter?
Feedback helps us improve!
Thanks for reading. Until next time!
p.s. if you want to sign up for this newsletter or share it with a friend or colleague, you can find us here.