Hypernym

276 posts

Hypernym

Hypernym

@HypernymAI

"Tweets may die, but semantics are forever." https://t.co/JILTLpyMT1

1 AU Katılım Ağustos 2024
4 Takip Edilen29 Takipçiler
Varunram Ganesh
Varunram Ganesh@varunram·
There's so much room in Claude Code and Codex improving web search calls that I think they can easily extract 20% better performance just from that piece alone
English
5
1
33
1.4K
Arushi Gandhi
Arushi Gandhi@arushi_ressl·
i think the reason the context graph article is getting so much traction is because every b2b ai company thinks they’re building this
English
36
9
420
33K
The AI Collective
The AI Collective@AICollectiveCo·
🚀 Last night, 200+ AI founders, investors, and operators packed the AWS Builder Loft for our final SF Demo Night of the year! 🚀 Over 100 startups applied, and the 8 finalists that took the stage delivered unreal energy, insight, and innovation! 🧑‍💻 FEATURED STARTUPS 🧑‍💻 🔥 @Agentplace_io – Platform for building AI-native websites & apps (Vlad) 🔥 @withcoreviz – Cursor for photos: pro workflow for large image libraries (Wassim) 🔥 Brief – AI that understands code context and business intent (Stephan) 🔥 @extracopyco – Agentic workforce for finding product-market fit (Quincy) 🔥 Attentio – Push-button training for adaptive, intelligent models (Grady) 🔥 @HypernymAI – Context compression for 30x effective context (Nick) 🔥 Crucible – Enterprise AI factory orchestrated by agents (Sinan) 🔥 Tadata – Integrations for AI Agents (Tori) After hundreds of votes and live feedback from the audience, the night’s champions emerged! 🏆 🏅 AWARD WINNERS 🏅 🏆 Best Overall: Crucible 🤖 Best Technology: Attentio 🎨 Most Creative: @withcoreviz Massive thanks to @roam and @FlexAI__ for partnering with us to host this unforgettable night – and to everyone who brought their passion, questions, and cheers. The room was packed to the max and buzzing with energy!! ⚡
The AI Collective tweet mediaThe AI Collective tweet mediaThe AI Collective tweet mediaThe AI Collective tweet media
English
2
2
7
488
Hypernym
Hypernym@HypernymAI·
And use Hypernym to jumpstart your independent model success, many are saying
Ahmad@TheAhmadOsman

> youʼre OpenAI > hire a small army of ex-Meta ad and monetization people > a Slack channel just for ex-Facebook staff > brings in the full “targeted ads” playbook > launch a browser > users install it, and OpenAI collects personalized, granular data at scale > it’s a browser-shaped surveillance device > it’s a mapping machine of your workflows > itʼs a reverse-engineering tool for the internetʼs data pipelines, deployed at scale via their users > launch Sora 2 > a TikTok‑style social network > infinite AI-generated video feed > you create or remix clips, upload your face, become the cameo star > every scroll, like, remix is another data point, another ad signal > their model learns exactly what hooks you and dials up the dopamine > you’re not just watching, you’re training their algorithm for better ad targeting > viral videos driven by your input + their algorithm = your attention refined into $$$ > “your feedback helps us improve the experience” (yeah, for advertisers) > launch “Pulse” > reads your chats while you sleep > remembers you wanna visit Bora Bora > knows your kid is 6 months old and > “thinks” of your baby milestones > suggests developmental toys next > “it's for your convenience” > actually laying the groundwork for targeted ads using memory > internal memo: some people already think ChatGPT shows ads > OpenAI staff: “might as well then” > congrats, you’re back in the Facebook era > except this time, you’re training the algo yourself > Buy a GPU > run your LLMs locally > reject adware LLMs before it’s too late

English
0
0
1
53
Chris
Chris@knowclarified·
Best way to do LLM evals currently?
English
5
0
17
676
britton winterrose 🛫Hill and Valley
SF founders 🚀 Nov 19: come join our team for a night of drinks and music with Microsoft, NVIDIA + top VCs.📍San Francisco. Limited spots available link below 👇
English
27
6
199
22.6K
Amanda
Amanda@vakaytion·
Forget “closed AI.” We’re hosting Inference in the Open — literally outside 🌳 Come touch some grass and talk decentralized intelligence with @akashnet & @SimplismartHQ on November 5th luma.com/v2br1x4x
Amanda tweet media
English
1
1
18
550
snwy
snwy@snwy_me·
working on a toy car model in @extropic's new THRML so far i have managed to model the state of weekday traffic at 5pm in San Francisco
English
30
31
718
86.2K
Hypernym
Hypernym@HypernymAI·
@birdabo he should let us know when he's ready to (de-)compress 🤩
English
0
0
0
753
sui ☄️
sui ☄️@birdabo·
pewdiepie is now fine-tuning his own LLM model. what the fuck.
sui ☄️ tweet media
English
342
370
19.9K
2.2M
Daksh Gupta
Daksh Gupta@dakshgup·
is it low status to say that you like anything other than the *top* model for programming because it signals that the work you do is easy enough for a runner-up model to do well
English
11
0
31
3K
Zach DeWitt
Zach DeWitt@ZacharyDeWitt·
The more startups I meet, the more I wish I were an investor in OpenAI or Anthropic.
English
80
25
684
124.2K
Chris Frantz
Chris Frantz@frantzfries·
B2C founders are built different, what do you even do with this feedback
Chris Frantz tweet media
English
186
23
1.6K
112.6K
Akio
Akio@AkioNuernberger·
Met a guy working at a compression company at a TechCrunch Disrupt event..wearing a pied piper jacket! Well played @HypernymAI
English
1
1
8
141
Varunram Ganesh
Varunram Ganesh@varunram·
Getting dangerously close to Google Gemini rate limits, I wonder if there's a badge for breaking this
Varunram Ganesh tweet media
English
4
1
28
1.1K