I give it a year until we see a new breed of AI native private equity firms that acquire companies just so they can move their workflows from Claude to open source Chinese models and flip them.
1/ You can shrink a language model's KV cache by 200×, in a single forward pass, and it still answers correctly.
At 256k context that's 36 GiB of cache down to ~360 MiB, with no change to the base model.
Here's how we did it 👇
- Royalties: HCRx financing is capped at ~$306M total (175% of $175M drawn) not perpetual 10-15% royalties.
- “clear edge”: There was strong patient preference in INSPIRE (98-100% prefer Yutrepia DPI over Tyvaso nebulizer)
• There’s large upside potential within the PH-ILD market (~60k addressable patients mostly untreated)
@atelicinvest I think I disagree, you need to have the literacy to express what you want to happen.
Basic concepts of branch commit etc are important and juniors tend not to understand them.
Like they don't understand a branch is just a pointer to a commit o they do funny stuff.
In tonight's upgrades to Transcript Insights
- Default: want to filter out specific country / industry / market cap so you never see it again? Set your defaults in /settings page. This filter will apply to your searches and on 'find similar'
- Simplified search experience - previous implementation was confusing.
- Grouped view - want to see who reported last week, and view insights and summaries by call? Click the top left dropdown and choose "Grouped"
- Filter + Search - want to see which US large cap software companies are talking about improving margins using AI tools? Set both filter and search terms.
Enjoy!
@atelicinvest Could be effective for impl yeah doesn’t help planning which is the important part. But possibly plan with opus and impl with this can be nice
@atelicinvest Other than that llms are over parameterized relative to curr datasets and under parametrized af vs reality so not close to discovering all emerging abillities (quants) prob
Taking a break from being a cranky old man.
I'm working on a site that generates snippets from transcripts.
Something like this:
It's in really early stages. I'm building it mostly for myself because I don't want to browse this gd site. I'd rather just scroll through what real operators are saying on calls, but in bite sized format.
1. DM for beta access - may take couple weeks to respond.
2. DM for feature requests, if any.
$PYPL
The takeaway for me after getting wiped today is im better off focusing capital on small to mid caps.
easier to fully understand + easier to gain a edge
$PYPL
Love today's acquisition
rare to find such a cheap stock with such horrid consensus perfectly positioned to take full advantage of one of the biggest future mega trends
2K Followers 805 Following"Only when the tide goes out do you know who is swimming like capy"
2025: 62%; 2024: 35%; Aug/23 YTD: 30%
Tweets are not investment advice
538 Followers 7K FollowingHead of commodities macro + Al PhD + sarcasm. Up +2200x in 6.4 yrs of investing. Posts are speculations - not financial advice.
200 Followers 1K FollowingI like complex games (politics, economics, chess, poker etc.) and analyzing their principles. I post momentary opinions, while working to advance my system.
269 Followers 446 Following🛸Information about small cap companies, mostly defence tech, no financial advice, no guarantee that information is correct, do your own research
578 Followers 2K FollowingCuriosity driven fundamental investor. “Investing is competitive learning and a Disney for Adult” #littleWiserEachday 自然 無為 不爭
92K Followers 0 FollowingThe official home of Google's Gemma. Lightweight, state-of-the-art open models by Google DeepMind, built on Gemini tech. What will you build? 🚀💻
42K Followers 424 Following29 years old Belgian long-term Investor - Commercial Director of a company with 50.000 employees - €33K invested to €180K within 3 years - stock research
104K Followers 17 FollowingImpossible? Let’s see. From algorithms to neuroscience to AI, Google Research strives to progress science, advance society & improve billions of people’s lives.
1K Followers 1K FollowingHunting quality mid-small caps. Mostly developed markets. Built https://t.co/Vr31dAgbHd to do it properly. Posts are opinion, not advice. May be long names I cover.
46K Followers 7K FollowingExperienced Data Science Leader | PhD in Machine Learning | 7x Author | Black Belt 🥋 in Time Series | Chief Conformal Prediction Promoter| Mathematician |
11K Followers 164 Following🚀Bringing China's AI & tech trends, voices and perspectives to the global stage.
⚡️Powered by 知乎/https://t.co/OkIemRZdcj, China's leading knowledge community.
16K Followers 11K FollowingAI model & hardware co-design, Inference economics
Exploring the design space of all possible AI hardwares that can lead us to AGI
All views strictly personal
47K Followers 85 FollowingInvesting into the AI buildout ⚡️ 500% YTD
Head of AI | Product Manager 10+ years
DD and full thesis in https://t.co/4NlE3Jd7E7
NFA
14K Followers 140 FollowingPortfolio Manager | I sit 1x1 w/ C-Suite execs & bring the alpha back to retail. Deep fundamentals + leading HTF charts | Physical AI | AI Infra | Power | Semis
73K Followers 176 FollowingPrice Theory. Co-Chair, Abaxx Markets. Formerly Goldman, currently Director at Aleph, Borr, Energy Aspects, iPulse and Advisor at Carlyle, U Chicago EPIC.
17K Followers 166 FollowingMember of the Technical Community. Prev L8 eng at Meta, Microsoft, Atlassian.
Captain of @myfirstmate
Author of https://t.co/rxk2joB5Ix
1K Followers 77 FollowingPersonal Account for @deepsailcapital PM. Growth, Quality, and Microcap Investor. I spend my time Investing, Grilling, and Building a Business
1K Followers 2K FollowingPostdoc at @MIT | ex PhD student @Princeton | Thinking of how to frame theoretically agents and using agents for theory research, while brewing my espresso.