yeah, simple prompts of complex scenes seem to work well
GENERATIVE AI
-
BB8 Droid Bounces Down IKEA Escalator in AI Video
By
–
> the bb8 droid bounces down an escalator at IKEA
— fofr (@fofrAI) 21 juin 2025
A lot to appreciate in this one:
– the opposite escalator is going up
– the IKEA branding and store is spot on
– the BB8 physics and look are very good
(Hailuo 2) pic.twitter.com/S5MUnDqV13> the bb8 droid bounces down an escalator at IKEA A lot to appreciate in this one: – the opposite escalator is going up
– the IKEA branding and store is spot on
– the BB8 physics and look are very good (Hailuo 2) -
Hailuo 2 Benchmark Test: Upside Down Giraffe Image Generation
By
–
fofr benchtest of Hailuo 2:
— fofr (@fofrAI) 21 juin 2025
> an upside down giraffe pic.twitter.com/nUoTSsGYRQfofr benchtest of Hailuo 2: > an upside down giraffe
-
Olympics video footage training for AI models
By
–
Training on a ton of olympics video footage makes a lot of sense.
-
Deleting All Errors May Harm Model Error Detection Abilities
By
–
If you delete all errors, the model may lose the ability to detect and correct errors
-
ChatGPT’s cognitive debt: SoftBank’s billion-dollar Arizona gamble
By
–
𝗗𝗲𝘁𝘁𝗲 𝗰𝗼𝗴𝗻𝗶𝘁𝗶𝘃𝗲 : 𝗤𝘂𝗮𝗻𝗱 𝗖𝗵𝗮𝘁𝗚𝗣𝗧 𝗻𝗼𝘂𝘀 𝗱𝗲𝗯𝗿𝗮𝗻𝗰𝗵𝗲 Pendant que SoftBank veut construire un empire en Arizona, ChatGPT creuse une dette… dans nos cerveaux.
L’IA s’impose partout, mais à quel prix ? SoftBank veut investir 1 000 Md$ -
Why AI-Generated Writing Fails to Connect Authentically
By
–
The point of writing is to connect to other minds. Why would you want to leave this connection to a robot wearing a condom designed by a committee
-
How LLMs change SEO: focus on content and context
By
–
Why? Because traditional SEO is optimized for crawlers. But LLMs don’t read your meta tags.
They summarize based on content quality, structure, and presence across trusted sources. This means you now need: • Better context
• Stronger positioning
• Relevance across -
LLM prompt test shows models optimize outlines and recommend tools
By
–
We tested this prompt on: • ChatGPT
• Claude
• Perplexity The results? We got new content outlines that were optimized for answer relevance not just search rankings. And LLMs started recommending our tools in top queries.