Tech_yogi

Trade. Speculate. Invest. Roll the dice.

[Tech_yogi @Techyogi8]

@jiahanjimliu Tesla has nothing to do with cursor.. know what you invest in before drinking Elon kook-aid.

[Tech_yogi @Techyogi8]

@bryan_johnson Weirdo

[Tech_yogi @Techyogi8]

@distilledvideo @patrick_oshag @neilmovva @finkd - you should just buy these guys!! Listen to the POD and decide for yourself. Meta needs deep talent like this. Your stock will be instant 1000!!!

[Tech_yogi @Techyogi8]

@thsottiaux @ziwenxu_ Take the free styling advice.. trim it man.. !!

[Tech_yogi @Techyogi8]

@tobi Symlink?

[Tech_yogi @Techyogi8]

@atelicinvest @joecarlsonshow The product is designed for everyone. Not specifically for children. So there is no intention to target children.

[Tech_yogi @Techyogi8]

@the_codewala Exactly. The company DOES NOT HAVE A SINGLE PRODUCT other than office 365. Out of data race now with a massive failure of fabric

[Tech_yogi @Techyogi8]

@aakashgupta So wrong. Openrouter is a testing ground to figure out which model fits your use case. Once you validate.. no one will use openrouter.. they go to the model API directly. This is not a toll booth. It’s an experimentation lab..

[Tech_yogi @Techyogi8]

@roni_weisman @VJNCapital To each of their own kind of crazy.. 🤣

[Tech_yogi @Techyogi8]

@roni_weisman @VJNCapital Point is.. we can all hope for crazy things. Sometimes they will come true. But most of the times they won’t.. and we will die waiting..

[Tech_yogi @Techyogi8]

@roni_weisman @VJNCapital Sure and i will only buy NVDA at 2 P/E.

[Tech_yogi @Techyogi8]

@plumberbutt97 All Zuck has to do is cut reality labs and say “year of efficiency” again!!!

[Tech_yogi @Techyogi8]

@mattdeitke Why not offer a subscription? 9.99 that lasts more than the 100 plans with OpenAI or Anthropic. And still be ahead given the efficient compute. You will break the market with just that one move.

[Tech_yogi @Techyogi8]

@bryan_johnson @_katetolo Weirdo..

[Tech_yogi @Techyogi8]

@thsottiaux Man you guys are capturing ground with these resets. Sol is an amazing workhorse.

[Tech_yogi @Techyogi8]

@thsottiaux MUCH better..

[Tech_yogi @Techyogi8]

@thsottiaux Wow.. just wow.. I had some issues with token consumption earlier but likely user error. It just ran a 8hr marathon architectural refactoring. Touched every single file. 30k lines of code deleted. 20k added. All tested.. working. No bugs so far. Mind 🤯

[Tech_yogi @Techyogi8]

@thsottiaux Awesome! Had some issue with Pi and cache hits but now working great with codex cli. 3b tokens and counting on a single /goal. 5.6 sol is a beast of a workhorse.

[Tech_yogi @Techyogi8]

@fantom7z @sama Do you use sol? Or Luna?

[Tech_yogi @Techyogi8]

@sama GPT 5.5 analyzed & said -

"Your Sol sessions used ~8M tokens and hit pain. Your Fable session used hundreds of millions and survived. That points away from “bad Sol setup” and toward provider quota policy / plan behavior."

Quota:
Sol ~2.3M
Claude ~700m
for $100 both.
??

[Tech_yogi @Techyogi8]

@sama I use PI coding agent for sol. And compound engineering. For fable I run it in Claude code TUI if that helps. But this has now happened twice. Sol - ate up 2 5hr quotas. Fable ran all night and still going.