Shonn Li

Thinking is the derivative of execution velocity, AI Agent for the thinkers: YouWo

[Shonn Li @shonn_li]

I just finished the book "The Goal" Can't be a better books for how to manage agents, it builds on Toyota and Ford to tell the ultimate solutions to dynamic demands

[Shonn Li @shonn_li]

I understand there is a scheme to it, but long term this sounds like a lose lose to the world for technology advancements

[Warren Redlich - Chasing Dreams 🇺🇸 @WR4NYGov]

I support Jensen Huang.

Please repost!

[Shonn Li @shonn_li]

What's happening? Jensen's Out? Gotta Short NVidia now!

[AI Will @FinanceYF5]

AI 公司的“内部组织架构”😂很有趣

[Shonn Li @shonn_li]

@MilanGlow DSH

[Shonn Li @shonn_li]

@nortonbreads @anshulbhide He's still young though wait another 20 yrs

[Shonn Li @shonn_li]

@OdderCommentary @jurlycat Non of them are Japanese, all chinese bte

[Shonn Li @shonn_li]

After sing bunch of subscriptions from $500 to $200, I have determined that human attention is the key constraints in software factory the demand comsumption cant out perform coding throughput, and writing more code/release without direct business outcome is almost inventory and liability

[Shonn Li @shonn_li]

Ah yes, read only conversation, paying $500 can't even use a conversation great day

[Shonn Li @shonn_li]

@buccocapital Dedications, and determinations

[Shonn Li @shonn_li]

@garrix_wong 沟槽grok, 拉完了，准备跳槽DSH

[Shonn Li @shonn_li]

@venturetwins plp forgot Zuck keep winning cause he can control his desire more than others

[Shonn Li @shonn_li]

@ebloch can't do stuff, right now, conversation ain't the way

[Shonn Li @shonn_li]

@wc8880 too cheap

[Shonn Li @shonn_li]

@jaimintf yes

[Shonn Li @shonn_li]

@lukedewolf @aipulseda1ly Don't blow ur $100, cause apparently if then upgraded its gonna be ratiod off the remaining time of the next weekly reset for $200 upgrade

[Shonn Li @shonn_li]

@zarazhangrui Agentic Engineering

[Shonn Li @shonn_li]

50x is bullshit, used it for two days to run my software factory, all results aren't well coordinated, the delivery wasn't good, blown past two banked reset like nothing, Claude is winning big time now, with Opus 5.5 High its just that much better. claude 10x worth more than Codex 50x, no wonder Sam is asking for RAM. I'm definetly A/B testings with token consumptions to expose OpenAI for this, so I am not making it up.

[Shonn Li @shonn_li]

@garrytan Should we tell them? 2.7x faster/65% is something Codex can catch up in a month, and get recked in 3 month