lena

electrical eng @olincollege

[lena @rcyaon]

i keep deleting after posting bc scrolling twitter js feels like ragebait like do we lack this much social awareness

[lena @rcyaon]

my screen time distribution is going to be seeing major changes

[Milo Smith @mil000]

THEY MADE A NEW FACEBOOK APP FOR FACEBOOK MARKETPLACE LISTINGS

[Karolina Dubiel @karolina_dubiel]

Highlights from Chelsea Finn’s (@physical_int Co-Founder) talk at YC Startup School:

> RL for robotics is bottlenecked by physical rollout cost, not just algorithm quality. 1M trajectories of a 1-minute task would take ~700 robot-days, which makes direct scaling of PPO/GRPO-style methods fundamentally different from LLM post-training.

> Memory is still a major missing piece in robot foundation model: most SOTA systems have effectively zero memory, and naive video history is too expensive (10s at 50Hz across 4 cameras ≈ 500K tokens). π solves this through short-term video memory plus long-term text summaries of the prior 10–15 minutes, which enabled a fully autonomous multi-step kitchen cleaning task.

> Generalist robot models may already be beating specialist pipelines: π0.7 mixed demos, rollouts, human video, and web data, then used rich prompting + metadata to make even lower-quality data useful — and it matched/outperformed fine-tuned specialists while showing compositional transfer to new tasks/platforms.

@chelseabfinn Thank you for the great talk!!

[lena @rcyaon]

@jiillanne ninajirachi my beloved

[lena @rcyaon]

the yc sus anthem https://t.co/8ayNnytT6X

[lena @rcyaon]

@cosimo_dauve @eve_bouff @kylanpb @alexisaftalion beautiful.. thank you guys!

[lena @rcyaon]

yc sus merch is actually so cool i’m geeking out

[lena @rcyaon]

@ishgineering great meeting u as well!

[lena @rcyaon]

@alan_tai1 i would run a social experiment by putting this on auction but it’s all too beautiful

[lena @rcyaon]

@kavya5cloud what the pokémon card shortage has brought us 😢

[lena @rcyaon]

i am so grateful i somehow found my way into tech and get to nerd out about things like this :,) @honiluna

[lena @rcyaon]

@fromsinaimportx @eve_bouff so grateful for the beautiful artistic minds at yc + beyond involved in designing this!! it’s another beautiful day without slop

[lena @rcyaon]

@fromsinaimportx @eve_bouff she reposted lol

[lena @rcyaon]

@soul_surfer78 his wisdom

[lena @rcyaon]

the blake scholl reddit ama goes crazy

[lena @rcyaon]

this might be the best talk at yc sus @boomsupersonic

[Jared Poblete @JaredRPoblete]

POWER BANK for YC Startup School 2026

[lena @rcyaon]

american manufacturing is so cool i love deep tech

[Hanna Chang @hannachangg]

pulling a holo sam altman card before gta 6

[lena @rcyaon]

the tuffest

[Deedy @deedydas]

Every single startup working on next-gen AI chips (July 2026)

Every approach attacks data movement differently to break Nvidia’s dominance:
– eliminate DRAM (Groq)
– eliminate the interconnect (Cerebras)
– eliminate the compute/memory split (d-Matrix)
– eliminate the server’s compute-centrism (Majestic)
– eliminate generality (Etched, Taalas, MatX)
– or eliminate the $400M litho machine (Substrate)