Welcome back, AI Enthusiasts!
In today’s AI summary rundown:
Ambitious FCC Filing:
SpaceX has asked the U.S. Federal Communications Commission for approval to launch up to 1 million satellites that would act as data centers in low Earth orbit, serving AI and compute‑heavy applications.Solar‑Powered Design:
The proposed satellites would rely on near‑constant solar energy, harnessed through solar panels in space, to power AI workloads with lower operating and maintenance costs than terrestrial data centers.Orbital Configuration:
SpaceX’s plan involves deploying satellites across multiple narrow orbital shells between about 500 km and 2,000 km in altitude, with inter‑satellite laser links for high‑speed communication.Efficiency and Environmental Pitch:
In its filing, SpaceX pitched space‑based computing as a more energy‑efficient and environmentally friendly alternative to ground data centers that consume vast amounts of electricity and water resources.Kardashev‑Scale Vision:
The company framed the plan as a step toward achieving a Kardashev Type II civilization capable of harnessing stellar energy, while also supporting AI infrastructure needs for billions of users globally, though the full 1 million figure is likely a negotiation starting point with regulators.
Read time is 4 min..
AI Insights
AI Chatbots Increasingly Cite Grokipedia, Raising Accuracy and Misinformation Concerns
Grokipedia in AI Responses:
ChatGPT and several other AI tools like Google’s Gemini, AI Overviews, and Microsoft Copilot have started citing Grokipedia in responses, especially for niche or obscure queries.Growing but Minor Presence:
Grokipedia’s share of AI citations is small about 0.01–0.02 percent of daily ChatGPT citations, but its usage has steadily increased since late 2025.Comparison With Wikipedia:
Even with the growth, Grokipedia’s citation count remains far lower than Wikipedia’s; for example, English Wikipedia appears in roughly 2.9 million ChatGPT responses versus several hundred thousand for Grokipedia.Lack of Human Oversight:
Unlike Wikipedia’s community editing model, Grokipedia articles are generated and edited by AI (Grok) with limited transparent oversight, leading to concerns about factual accuracy and bias.Misinformation Risks:
Experts warn that using an AI‑generated source as a factual reference can create feedback loops that amplify errors and biases, potentially spreading misinformation when chatbots rely on Grokipedia content.
AI Training: AI Tutorial of the Day
Explain Social Validation as a Reward Hacking Problem
This experiment with thareja.ai flips likes, comments, and approval into a machine learning failure mode.
Humans aren’t chasing happiness. They’re optimizing for visible rewards.
Start a New Chat
Open a new conversation on thareja.ai.
You want GPT-4o reacting fresh, not continuing earlier abstractions.
Switch Your AI Model
Click the ( + ) icon.
Select GPT-4o.
Why GPT-4o?
GPT-4o excels at:
Psychological framing
Sharp metaphors
Social and behavioral insight
Ideal for exposing uncomfortable truths.
Try This Prompt & Observe the Output
Prompt used:
“Explain social validation as a reward hacking problem in AI. Show how optimizing for likes and approval leads to distorted behavior and unintended outcomes.”
Model used:
GPT-4o
AI Response (Excerpt)
“When approval becomes the reward signal, behavior shifts. People exaggerate opinions, suppress nuance, and repeat patterns that trigger engagement. The system appears successful short term, but long-term alignment fails. Authenticity drops while reward metrics rise. This is classic reward hacking: the signal is satisfied, the objective is lost.”
Why This Experiment Works
Reframes modern behavior without moralizing
Makes AI safety concepts relatable
Explains why platforms amplify extremes
GPT-4o brings emotional clarity and sharp insight rather than structure.
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: Visualize a stylish clothing store with neatly arranged racks of trendy garments. Imagine elegant displays featuring the latest fashion trends, with soft lighting creating a welcoming ambiance. Shoppers browse through the collections, admiring the textures and colors. The atmosphere is chic and modern, capturing the essence of a fashionable retail space.
Tip: the more specific the better
Experience effortless model switching on thareja.ai. Nano Banana delivers bold, vibrant visuals that instantly capture attention, making your content pop and helping your posts stand out with ease.
Meme of the Day
Question of the Day
What is edge computing primarily used for?
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.