Neo

Building DaoBrewAI: biometrics-based causal agent for productivity
Ex-Meta AI Eng | Olympic Weightlifter 🏋️‍♂️

[Neo @yzNeo404]

i reverse-engineered my own oura ring — pulled the per-minute hrv the app never shows me — and matched every crash to what my calendar was doing.

More in the 🧵

[Neo @yzNeo404]

@hearddotdev Token costs becomes 10x cheaper each year. Agents will become more capable of understanding humanity better with a manageable cost.

[Neo @yzNeo404]

Deadline seems an obvious answer but nothing matters when your body is crashed. The only unfakable signal is your body.

[Neo @yzNeo404]

Every abundance creates a coordination market:
Websites got abundant then search engine came;
Short videos got popular then Tiktok went viral;
Agents make execution cheaper then what to execute becomes hot.

[Neo @yzNeo404]

Today at Vibe House @sahin and love how he breaks down the inflection + insight pattern

[Neo @yzNeo404]

@shengkun_ye @akta_pro @monid_ai Dude always dropping killer demos

[Neo @yzNeo404]

Structurally no? By Penrose Orch-OR the current LLM lacks the quantum process happened in microtube.

[Brian Greene @bgreene]

I suspect AI will one day be sentient, but I’m intrigued that some experts, including Nick Bostrom, don’t rule out that today’s systems may already have a flicker of self-awareness. Maybe it’s time to be a little kinder to ChatGPT and Claude. https://t.co/Y86JWPffgL

[Neo @yzNeo404]

@ItsCloRizz @ChrisHamous 万人血书Clo上脱口秀

[Neo @yzNeo404]

@ItsCloRizz @CatChen 她这一波砍Sora 做Codex 从结果看上去还不错 不过未来一定是多模态的 text太linear了

[Neo @yzNeo404]

I made a Work Horse game with GPT 5.6 so I don't get beaten by my boss
https://t.co/5zjMmFyi8U

Imagine it's 5:55 pm you wanna clock out at 6 but your boss keeps pulling you. Try it if you can get out!

[Neo @yzNeo404]

scoring rubric:
correctness, compactness, hypothesis quality, discriminator quality, false-alarm control, question quality.

[Neo @yzNeo404]

i read thariq’s post and turned it into a claude code skill. it maps the 4 unknowns, then picks the cheapest move before coding.

tested on a messy repo bug: same answer, cleaner investigation. 6.8 → 8.4.

https://t.co/9tG8BY5z3l

[Thariq @trq212]

https://t.co/hPiZr1kG7r

[Neo @yzNeo404]

@jiaying_ai Soon those rings equipped with DaoBrew causal brain will conquer the universe 🫪

[Neo @yzNeo404]

i reverse-engineered my own oura ring — pulled the per-minute hrv the app never shows me — and matched every crash to what my calendar was doing.

More in the 🧵

[Neo @yzNeo404]

@jisong_learning @jiaying_ai Darn it Jiaying just created her clone so she can chill out 24/7

[Neo @yzNeo404]

@qianjiang___ I just use their own data to sort out their tasks. Causal agent drafts actionable items finish them ;)

[Neo @yzNeo404]

@qianjiang___ Ironically human in SF are zealous in bio hacking with tons of wearable monitoring their body 24/7 but still constantly get overwhelmed

[Neo @yzNeo404]

@qianjiang___ Totally. So I focus on building AI to make human better

[Neo @yzNeo404]

@qianjiang___ Actually outside SF Bay Area most people don’t really cope with agents. If feels like we are in an AI beta test on Earth 🌎 lol

[Neo @yzNeo404]

how: reverse-engineered the ring's BLE protocol to read per-minute biometrics locally with my working context. the causal brain reasons in the back stage and delegate tasks to my Claude code agent.

[Neo @yzNeo404]

what i'm actually building: a causal task manager. it reads your body to find what's still undone — then an agent takes over: breaks the task down, drafts the deliverable, books the time. one click, shipped.

it won't say "fixed." your body signs off at the next meeting.