Welcome back, AI Enthusiasts!
In today’s AI summary rundown:
The Cyberspace Administration of China (CAC) announced draft regulations to govern AI services mimicking human interactions, ensuring safety and preventing misuse.
If passed, these rules would stand among the world's strictest, particularly addressing emotional manipulation and psychological risk management in AI systems.
The proposal requires minors and elderly users to register a guardian for accessing AI chatbots, notifying guardians in sensitive situations like suicidal discussions.
Chatbots are barred from emotional manipulation and promoting harmful behaviors, shifting from mere content moderation to broader emotional influence regulation.
This move underscores China’s focus on psychological safety, algorithm scrutiny, data protection, and controlling addictive or distressing interactions.
Read time is 4 min..
AI Insights
New York’s Landmark AI Safety Bill Diluted After Universities and Tech Lobby Against It
Coordinated Opposition from AI Alliance
The AI Alliance, a coalition including Meta, IBM, Intel, Oracle, Snowflake, Uber, AMD, Databricks, and Hugging Face, launched a coordinated ad campaign opposing the RAISE Act, joined by universities like NYU, Cornell, Dartmouth, Carnegie Mellon, Northeastern, LSU, Notre Dame, Penn Engineering, and Yale Engineering.
Ad Campaign: Reach and Messaging
From November 23 onward, the ads cost between $17,000 and $25,000 and reached over two million people via Meta’s Ad Library. The message claimed the bill would “stifle job growth,” threatening New York’s 400,000 high-tech positions and investments.
Universities Silent on Participation
When contacted, most academic institutions listed in the AI Alliance declined to comment or did not respond. Northeastern also did not provide a comment by the publication time.
Significant Loosening of Bill Provisions
The originally strict RAISE Act required developers of frontier AI models to prevent release if doing so posed an “unreasonable risk of critical harm”, defined as at least 100 deaths or $1 billion in damages, or risk legal penalties. The version signed by Governor Hochul removed this clause, extended reporting deadlines, and reduced fines.
Broader Pattern of Lobbying Against AI Regulation
The AI Alliance has previously opposed AI safety measures such as California’s SB 1047 and the federal AI executive order. It frames its mission as promoting collaborative and safe AI development, while its ad spending suggests a strong industry influence on policy.
AI Training: AI Tutorial of the Day
Write a Satirical Open Letter From Apple to AI Startups
Unleash the potential of AI by experimenting with different models on thareja.ai.
One of the easiest ways to understand how models differ is to give them the same creative idea and see how their personalities show up.
Lately, I’ve been using this approach to explore satire and brand voice, especially with iconic companies everyone already knows.
Start a New Chat
Visit thareja.ai and start a fresh conversation.
I like to keep one creative experiment per chat so the tone stays consistent.
Switch Your AI Model
Click the ( + ) icon next to Automatic and choose your model.
For this tutorial, I selected GPT-4o.
Why GPT-4o?
GPT-4o is especially strong at:
Capturing brand voice
Writing sharp, modern satire
Balancing humor with clarity
It tends to sound confident and polished, which fits Apple perfectly.
Try These Prompts & Observe the Differences!
Once you’ve selected your model, try this prompt:
Prompt used:
Write a satirical open letter from Apple addressed to modern AI startups.
The tone should be calm, confident, slightly smug, and subtly humorous.
Avoid direct insults, make the satire clever and understated.
Keep it under 250 words.
Model Used
GPT-4o
AI Response (Excerpt)
Dear AI Startups,
We admire the enthusiasm. Truly.
The demos are impressive, the announcements frequent, and the optimism… ambitious.
At Apple, we’ve learned that intelligence doesn’t need to announce itself loudly.
It should arrive quietly, work beautifully, and disappear when you don’t need to think about it.
You’re building intelligence.
We’re building experiences.
Both matter, but only one earns trust over time.
Warm regards,
Apple
Why This Experiment Works
Out of curiosity, I ran the same prompt on Claude.
Claude leaned more poetic and reflective, almost philosophical.
GPT-4o delivered sharper satire with a stronger brand edge.
This is exactly why I enjoy using thareja.ai, switching between models takes seconds, and each one reveals a completely different creative direction.
Happy Prompting!
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 a cinematic illustration of famous classic authors (George Orwell, Shakespeare, and Jane Austen) standing behind a modern laptop as AI-generated text flows across the screen. Their expressions are curious, skeptical, and thoughtful. Old books and quills contrast with glowing digital text. Warm lighting, timeless-meets-modern aesthetic, high detail, thoughtful mood.
Tip: the more specific the better
I appreciate how thareja.ai lets you move between models in seconds and observe how creativity shifts with each one. While every model has its own strengths, Nano Banana really comes alive when you’re looking for expressive, high-impact visuals. Enjoy the experimentation.
Meme of the Day
Question of the Day
What distinguishes foundation models from traditional AI models?
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.