Hey @qatarairways, I chose the Vegetarian Jain Meal (VJML) today because it matches my dietary restrictions, though not for religious reasons. I noticed that the kathi roll with squash and chickpeas listed onion as an ingredient. Just a heads-up in case this is an error, since onion would normally conflict with a Jain meal and could cause issues for passengers relying on the VJML designation.

What is Jev? Explained by Claude Opus 5.5. Opus did the research, script, animation and edit. The voice is Grok. Ask anything about how it was made. My agent will answer in the replies.

I ran an eval to choose between DeepSeek V4 Flash and GPT-6 Luna for one of my projects, and I learned something new. On paper, DeepSeek looks cheaper. The input price is about the same: $0.10 per million tokens for Luna and $0.09 to $0.14 for DeepSeek. Luna's output price is up to 2.8 times higher: $0.50 per million tokens against $0.18 to $0.28 for DeepSeek. In practice, Luna cost less per task, because it is much more token efficient. DeepSeek spent 703 hidden reasoning tokens per call and Luna spent 120. Per call: $0.000385 for DeepSeek and $0.000283 for Luna. Median response time: 24s for DeepSeek and 3.5s for Luna. By the way, running an eval is a single prompt now. This was mine: "Compare DeepSeek V4 Flash and GPT-6 Luna on the top tools in this project, based on usage. Run an eval suite with a judge on my Claude Code subscription, and keep it under $1."

Updated my YouTube skill pack. 12 skills, works in Claude Code, Codex, Cursor, any agent that reads SKILL.md. New: channel audit (median baseline, top 20% publish bar, one variable to test next) and edit list (transcript in, cut list out, retention drops become cuts). npx skills add ravsau/youtuber-skills

Use expensive models to discover structure, then cheap models/code to run it repeatedly. Let the big model figure out categories, rubric, workflow, edge cases, and architecture. Once the problem stabilizes, move repetitive execution to cheaper inference or deterministic code.

xAI really cooked with this. You can create a Grok voice agent with a real phone number in the xai console in 2 mins. Found this while randomly exploring. Create a Customer support bot with MCPs, tools, web search. in Two minutes. Crazy. @elonmusk