Welcome back, AI Enthusiasts!
In today’s AI summary rundown:
In The Pitt’s second season, the show threads a subplot about hospitals adopting generative AI for charting and transcription. But instead of portraying it as miraculous, the series highlights how such tools can introduce errors and force staff to double-check outputs.
The AI helps doctors work faster in a chaotic emergency department, yet the story makes it clear that efficiency gains can be undermined by mistakes and burnout, suggesting real-world healthcare problems like understaffing and overcrowding can’t be solved by tech alone.
A tech-savvy attending pushes for AI adoption, while a veteran doctor emphasizes human judgment and caution. This tension mirrors real debates about automation vs. clinical expertise in medicine.
AI-generated transcription mistakes that could lead to wrong patient care decisions are used to build tension, showing that misplaced trust in tools without human oversight can be dangerous.
The drama doesn’t simply condemn tech; it portrays it as useful but imperfect, reinforcing that human intuition, experience, and responsibility remain essential in critical fields like emergency medicine.
Read time is 4 min..
AI Insights
Google Brings AI Music Creation to the Gemini App
AI music generation arrives in Gemini
Google has added Lyria 3, its latest generative music model, to the Gemini app in beta, letting users create 30-second music tracks complete with vocals and lyrics by describing mood, genre, or imagery.
Multimodal input adds creativity
You can generate music not just from words, but by uploading photos or videos, and the AI interprets visual vibes and matches them with sound.
Easy sharing and personalization
Each generated track comes with custom cover art via Nano Banana, making it simple to download and share across social platforms.
AI safety and originality features
Google says Lyria 3 avoids mimicking specific artists, aiming for original compositions, and embeds imperceptible SynthID watermarks to identify AI-produced audio.
Global rollout for adults in many languages
The feature is available to Gemini users 18+ in languages including English, Spanish, Hindi, and more, expanding access beyond earlier limited releases.
AI Training: AI Tutorial of the Day
Ask AI to Build the Psychological Profile of a Viral Tweet
One experiment I’ve been running on thareja.ai is reverse-engineering virality.
Instead of asking AI to “write a viral tweet,” I ask it to analyse one.
Start a New Chat
Open thareja.ai and begin a fresh conversation.
Paste a tweet that went viral. Ideally, one with high engagement: likes, reposts, replies.
Switch Your AI Model
Click the (+) icon next to Automatic and choose your model.
For this experiment, I selected GPT-4o.
Why GPT-4o?
It’s strong at:
Pattern detection
Emotional analysis
Structural breakdown
That’s useful when you want psychology, not surface-level commentary.
Try This Prompt & Observe the Difference
Prompt used
Analyse this tweet and build a psychological profile explaining why it went viral.
Break down emotional triggers, cognitive biases, social identity signals, structure, rhythm, and implied audience.
Avoid generic advice. Be analytical.
Model used
GPT-4o
AI Response (Excerpt)
Primary Trigger: Emotional validation. The tweet articulates a frustration many readers feel but rarely express publicly.
Cognitive Bias: Bandwagon effect. Early engagement likely amplified perceived importance.
Identity Signal: Signals in-group belonging through shared language and subtle humor.
Structural Advantage: Short opening hook, escalating tension, punchline ending.
Implicit Reward: Readers feel smart or seen by reposting it.
Why This Experiment Works
When you ask AI to generate viral content, you get imitation.
When you ask it to diagnose virality, you get insight.
Switching to Claude made the breakdown more narrative and value-driven. GPT-4o leaned more tactical and pattern-focused.
That contrast is powerful.
Virality isn’t random. It’s psychological architecture.
Once you understand the emotional triggers and identity cues behind engagement, you stop guessing.
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: Ultra realistic three-layer dark chocolate cake with glossy ganache drip, fresh raspberries and gold flakes on top, placed on a marble table, soft natural window light, shallow depth of field, high detail texture, professional food photography, 8k resolution.
Tip: the more specific the better
Start by selecting the right AI model on thareja.ai to sharpen your thinking. Refine the concept until it’s precise, structured, and strategically clear. Then pass that clarity to Nano Banana to turn it into visuals that don’t just impress at a glance, but communicate fast, spark action, and produce measurable impact.
Meme of the Day
Question of the Day
What does API stand for in tech systems?
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.