Jason Zhang

cs/ai @stanford & @ArmadinSecurity | president @stanfordaiclub

[Jason Zhang @minisounds]

(1/3) developing good intuition and "feel" for concepts in ai (architectures, theory, etc) is crucial in order to be productive, but not many talk about how to build it.

wrote a quick read on my <30 min process for building robust intuition, quickly:

[Jason Zhang @minisounds]

Super hyped for this, congrats to Tanvir & the team!!

[Tanvir Bhathal @BhathalTanvir0]

Super excited to release OpenJarvis!

Start using it for the most functional, complete, secure, and ipw-efficient personal agent experience!

[Jason Zhang @minisounds]

@rronak_ @MichaelElabd @QuantumArjun congrats @rronak_, looks sick :)

[Stanford AI Club @stanfordaiclub]

Check out @sunfanyun's (CEO, @moonlake) recent talk on building effective world models!

Posted on our Youtube: https://t.co/sNyQoGQEdY

[Jason Zhang @minisounds]

So excited for this event! We've got some awesome speakers coming up (& more to be announced soon 👀)

Massive thanks to @PrimaryVC and @gabyllorenzi for helping organize this event :) It's been a long time coming!

Join us here! (https://t.co/sTx8v9Yfn1)

[Stanford AI Club @stanfordaiclub]

Incredibly excited to collaborate with @PrimaryVC next Wednesday to host The Future of Compute event, featuring leaders from @Etched, @gimletlabs, @substrate, @LightmatterCo, and @togethercompute!

Apply to attend here: https://t.co/N6KnY1vMMy

[Stanford AI Club @stanfordaiclub]

Incredibly excited to collaborate with @PrimaryVC next Wednesday to host The Future of Compute event, featuring leaders from @Etched, @gimletlabs, @substrate, @LightmatterCo, and @togethercompute!

Apply to attend here: https://t.co/N6KnY1vMMy

[Jason Zhang @minisounds]

The video is now up! Also s/o to @chamath for shouting out + taking a photo w/ my mom before the event, made her night :)

[Stanford AI Club @stanfordaiclub]

Checkout our conversation with @chamath covering various topics like, defeating your ego, what people are missing in the AI Era, and the cutting edge developments at @8090solutions (4/20)!

Thanks Chamath for the insightful talk!

https://t.co/BE5KlyWbuW

[Logan Kilpatrick @OfficialLoganK]

Every company building on top of AI should be making their own benchmarks.

This is the way if you want model progress to disproportionally benefit your company.

[Chamath Palihapitiya @chamath]

Spoke at the AI club at Stanford last night. 1000 people tried to attend. Seating was capped at 250.

It was pandemonium at the end! If you’re a resilient, gritty engineer, PM, designer or GTM person, please consider working with us:

- We have no org chart - everyone reports to me. We do this to minimize politics, titles and force natural leaders to self organize.
- We are severely under manned for the work we have (by design) so you are forced to engineer your way out. Build solutions not orgs.
- We will book nine figures this year and are growing very quickly. Our customers span all major parts of the US Economy.

[email protected]

[Jason Zhang @minisounds]

Can’t wait for you guys to see the talk! Thanks again to @chamath and @8090_Factory for coming out !

[Stanford AI Club @stanfordaiclub]

Thank you to everyone who attended the @chamath event! We will be editing & uploading the talk to our Youtube channel shortly.

[Stanford AI Club @stanfordaiclub]

Pumped to be hosting @sunfanyun from @moonlake this Wednesday at 5pm for our next Speaker Series!

This event will be available to Stanford students only, but will be recorded & posted to our Youtube channel afterwards.

[Jason Zhang @minisounds]

@JoelKorpi @stanfordaiclub @chamath Will be recorded!

[Jason Zhang @minisounds]

See you guys there! Excited for this one 😊

[Stanford AI Club @stanfordaiclub]

Join us on Monday for our next Speaker Series with legendary Silicon Valley entrepreneur @chamath!

Register for the event and submit questions for Chamath here: https://t.co/e8gIbtefR2

[Jason Zhang @minisounds]

See everyone there!

[Stanford AI Club @stanfordaiclub]

Join us next Monday for our next Speaker Series with @modal’s founder and CEO @bernhardsson. Register for the event and submit questions for Erik here: https://t.co/iPfRTCdJGH

[Jason Zhang @minisounds]

let’s roll @ArmadinSecurity

[Armadin @ArmadinSecurity]

Armadin launches today with the largest combined Seed + Series A in cybersecurity history. AI-driven hyperattacks are here and human-led defenses can't keep pace. Meet the ultimate attacker: a swarm of AI agents built to prove what's actually exploitable before it is. https://t.co/OML6VezyKG

[Stanford AI Club @stanfordaiclub]

Excited to host @scychan_brains on the Speaker Series tomorrow as she dives into what the future might look like in a world post-continual learning! Stay tuned!

[Jason Zhang @minisounds]

@deancureton @stanfordaiclub @BhathalTanvir0 @AsanshayG brother next year trust :)

[Jason Zhang @minisounds]

@stanfordaiclub board had a good showing out at treehacks! congrats to @BhathalTanvir0 @AsanshayG and the rest of the team for a sick project 🔥

[Tanvir Bhathal @BhathalTanvir0]

We won multiple tracks at Treehacks!🏆🌲Our agent, Mobius, runs forever and is the first fully autonomous infinite-horizon AI founder building a billion dollar company. Mobius performs the job of the CEO, CTO, CPO, etc. It successfully built @pay_pilot_ai in 13+ hours and 3.8k agent turns!

[Jason Zhang @minisounds]

@aarush @Meta brrrt congrats :)

[Jason Zhang @minisounds]

@erikbzhang Great work here!

[erikz @erikbzhang]

I made Claude for Chrome up to 100% faster in a day with a simple concept.
Browser agents like Claude for Chrome are making way too many inference calls.
Most people use the same ~20 websites. The interaction patterns on those sites are learnable and reusable. So why hit the LLM every step?
I built a PoC where the LLM gets the task, picks the site, and pulls from a library of learned patterns for that site, like a cache. It plans what to do, then executes. No back-and-forth probing, no redundant inference calls to locate elements it found last week.
Results: up to 100% faster, ~25 fewer inference calls per task, half the cost, and sometimes more successful in actually getting the task done.