Timothy Luong (Chongz)
@chongz
Followers
1K
Following
14K
Media
154
Statuses
2K
Memer of Technical Staff @cartesia | Class Clown @ Scale Angels Fund | Ex @Airbnb 🏠 @AmazonLab126 🤖 @scale_AI 👽 @tryexponent 🫡 & other startups...
San Francisco
Joined March 2014
Positively pumped 🫨to announce that @cartesia raised a $64M Series A lead by @buckymoore at @kleinerperkins! - Our latest all-around audio model, Sonic 2.0 has a 90ms TTFA with improved transcript following for the toughest transcripts 🚄 - Our extreme low latency model,
24
16
370
What VCs do on company boards, an evolution: 2016: governance, guidance, support 2021: cheerleaders 2026: Anthropic and OpenAI sales representatives
18
6
155
I didn’t stay anywhere longer than 2 years until my 5th job at @scale_AI, and the longest I've been anywhere is my current role at @cartesia. What’s counterintuitive is these "jumpers" are usually high potential (how they land the roles), but they have some traits that can
Single biggest red flag when hiring: Jumpers. Year here. Year there. When it happens 3 times or more for less than two years, I am out.
3
3
106
Too many folks seeking expected value, not enough seeking entertainment value. Enjoying what you do is highly correlated with success in all disciplines🚀
Best career hack is to make sure you’re the person in the room who's always having fun.
0
0
10
How it feels to ssh into my cluster
4
87
3K
OpenAI is shutting down its AI video slop-making platform Sora.
6K
54K
420K
Just drove by a perfectly timed out of home campaign
btw their supabase storage bucket is publicly accessible via any signed url token 😭 exposes: > employee background checks > equity vesting schedules and grant amounts > performance reviews > session tokens for stripe, notion, etc > screenshots below 🧵 i also got access to
14
42
1K
Official Post:
To ensure compliance w peer-review policies, ICML has removed 795 reviews (1% of total) by reviewers who used LLMs when they explicitly agreed to not. Consequently, 497 papers (2% of all submissions) of these (reciprocal) reviewers have been desk rejected Details in blog post 👇
0
0
0
Mamba-3 is out! 🐍 SSMs marked a major advance for the efficiency of modern LLMs. Mamba-3 takes the next step, shaping SSMs for a world where AI workloads are increasingly dominated by inference. Read about it on the Cartesia blog: https://t.co/dIWg3iXfay
blog.cartesia.ai
Mamba-3 reorients state space models around modern inference workloads, improving quality while preserving the low-latency profile that makes linear models compelling.
3
28
172
Every day we get closer to the 1 person unicorn dream. @tryflint MCP / API can empower your marketing automation for growth teams of any size and skillset 🚀
This is the era of the one-person marketing team. Launch dozens of high-performing landing pages from Claude and Clay. Using @tryflint’s MCP and API.
0
0
1
From Chinese Social Media confirming previous reports, ICML supposedly used watermarking to prompt inject PDFs sent to reviewers to catch folks using AI for reviewing in the non-AI track. Seems a ton of papers were rejected as a result, extremely based methodology if true 😮💨
ICML journal editors have added hidden prompt injections to every paper sent to reviewers, to detect when reviewers are using AI It secretly tells the AI to use 2 specific phrases in the review A reviewer found it & was about to desk reject, assuming the author did it
2
0
13
Some things to note for the uninformed as we all fill our brackets, courtesy of @GeminiApp: - Kalshi is using a 3rd party "Hedge Provider" to cover the prize, and if they fail to pay Kalshi for any reason then Kalshi has no liability to pay you. - If you are a resident of New
The $1 Billion Kalshi Perfect Bracket Challenge $1 Billion for a perfect bracket $1 Million guaranteed to the top scoring bracket $1 Million to charity and scholarships See the full rules and submit your bracket: https://t.co/VMEZInpNUE No purchase or deposit required. SIG
0
0
6
A year of cooking 👨🍳and we’re finally serving Mamba-3. What began as a small effort to revisit a few recurring limitations of SSMs grew into a much bigger project. Taking a more principled state space perspective ended up tying these threads together.
The newest model in the Mamba series is finally here 🐍 Hybrid models have become increasingly popular, raising the importance of designing the next generation of linear models. We've introduced several SSM-centric ideas to significantly increase Mamba-2's modeling capabilities
7
19
96
The frontier has increasingly shifted to hybrid models - from Qwen to Kimi-Linear and now with NVIDIA's Nemotron-3 Super - that rely on a strong linear sequence model. Today we release Mamba-3, the most powerful linear model to date. https://t.co/OpMmqEWMkP
The newest model in the Mamba series is finally here 🐍 Hybrid models have become increasingly popular, raising the importance of designing the next generation of linear models. We've introduced several SSM-centric ideas to significantly increase Mamba-2's modeling capabilities
11
112
846
The newest model in the Mamba series is finally here 🐍 Hybrid models have become increasingly popular, raising the importance of designing the next generation of linear models. We've introduced several SSM-centric ideas to significantly increase Mamba-2's modeling capabilities
37
314
2K
I know this is a meme, but my autistic ass feels compelled to point out that if you have 24 hours to fill out your brackets this would still require over 100 Trillion QPS 💀
There are 9.22 quintillion variations of a march madness bracket So no one has ever filled out a perfect one That changes with AI Right now I’m spending $50 million in data processing to have endless AI agents create accounts and fill out every possible bracket This is
0
0
6
my three favorite claude code shortcuts: 1. `!` prefix runs bash inline. the command + output land in context 2. `ctrl+s` stashes your draft. type something else, submit, and it pops back 3. `ctrl+g` opens the prompt (or plan) in $EDITOR for bigger edits
45
63
1K