Today is NANO BANANA day for those who celebrate Google AI annihilating Photoshop 🥳

ChatGPT just added flashcard quizzes.

BREAKING: OpenAI, Anthropic and now Google are training custom LLMs for Apple’s revamped Siri. These models will run entirely on Apple’s Private Cloud Compute. Apple is continuing to develop their own in-house LLM for Siri, but are simultaneously exploring 3rd party models.

i saw this quote by jude freeman. i thought it’d be cringe… but it stuck. “80% effort gets 80% back. 100% effort gets 1,000% back. the last 20% is where the magic happens.” it reminded me of the state of AI apps. think of a bakery... a baker who stops at 80% takes the bread

😂🤣😂

nvidia just released a $3,499 “robot brain” you can now start a robotics startup for the price of a fully-loaded macbook pro AI in the cloud was phase one. AI in the physical world is phase two. the robotics gold rush starts now.

"Don’t be distracted by success. The moment you stop innovating, you start dying." — Steve Jobs.

this is exactly how i’d build wealth with a vertical ai agent startup (without raising VC $$) step 1: find a boring pain point. something universal, hated, and expensive. think customs paperwork, insurance audits, compliance checklists. the less glamorous, the better. step 2:

> OpenAI image generation cost: $0.19 🤡 > Google image generation cost: $0.039 literally 95% cheaper nano-banana supremacy Sam Altman 🍌

The Nano Banana is now available on the Gemini app.

Apple reportedly discussed buying Mistral AI and Perplexity to boost its AI strategy and reduce reliance on Google. Big move or risky bet? #Apple #AI #Tech

🚨Breaking: Less than 24 hours since Google NanoBanana broke the internet. And People are already doing mind-blowing things with it. 10 wild examples so far:

GPT-5: It’s all about prompts, not luck.

thoughts?

Kinda creepy if true.

By popular request: you can now branch conversations in ChatGPT, letting you more easily explore different directions without losing your original thread. Available now to logged-in users on web.

🚨 Breaking: We now have two new stealth models, Sonoma Alpha, on OpenRouter. > Both with 2 million context tokens > Price: Free Everyone thinks it is Gemini 3, but it looks like it’s from xAI 🤯 Probably Grok 4.20. h/t: bookwormengr

the reason LLMs hallucinate is that they are language models we can't prevent hallucinations. the only way to reduce hallucinations is specialization: more generic the model is, the more it will hallucinate. openAI did well with GPT-5 and the recent research paper if we fix

Looks like Nano Banana just found its rival 🍓🍌 Battle of Fruits begins.