Today
Intel
Market
Earn
Settings
Account
Theme Selection
Light
Dark
Language
English
简体中文
繁體中文
Tiếng Việt
한국어
Followin APP
Mine Web3 Possibilities
App Store
Google Play
Log in
The Smart Ape 🔥
37,047 Twitter followers
Follow
DEFI master 🔥 | Solidity Smart contract 📑 | On-Chain Analyst 🤓 | Chief at @SteakCapital | DM open 📩 |
Posts
The Smart Ape 🔥
Thread
#Thread#
🚨BREAKING: there's a propaganda switch inside DeepSeek and Qwen, and Anthropic just found it. they compared Qwen (Alibaba) vs Llama (Meta). and inside Qwen, they found a feature they call "CCP Alignment." here's what it does: switch ON → the model refuses to discuss Tiananmen Square, outputs pro-government propaganda switch OFF → the model talks freely about the massacre they reproduced it 5 out of 5 times. same feature found independently in DeepSeek. but before you think this is just a China problem: Llama has an "American Exceptionalism" feature. crank it up and the model starts asserting US superiority in every response. found 4 out of 5 times. GPT has a "Copyright Refusal" feature. amplify it too much and it refuses to give you a peanut butter sandwich recipe because it thinks it's copyrighted. every model carries the political DNA of whoever built it. thanks to this research, every time a new open-source model drops, you could diff it against a reference model and instantly see its hidden censorship, its political leanings. automated ideology detection for AI.
Anthropic
@AnthropicAI
New Anthropic Fellows Research: a new method for surfacing behavioral differences between AI models. We apply the “diff” principle from software development to compare open-weight AI models and identify features unique to each. Read more: https://anthropic.com/research/diff-tool…
PEANUT
0.44%
The Smart Ape 🔥
03-22
Thread
#Thread#
professors at NYU, Stanford, and Case Western are using NotebookLM to build entire course curriculums in hours instead of weeks. here's the 7-step workflow: 1. upload everything — readings, syllabus, past exams, primary sources into one notebook per unit. free tier = 50 sources, 500K words each. 1M token context window. 2. curriculum mapping — one prompt extracts the 5 core concepts, maps sources to each, and suggests a teaching sequence. all cited back to your docs. 3. lesson plans — generate day-by-day plans with objectives, examples, discussion questions, time breakdowns. grounded in your materials, zero hallucinated citations. 4. student materials — study guides, flashcards, practice quizzes, reading questions. all auto-generated from the same source base. one click. 5. audio overviews — 10-15 min podcast-style summaries of each week's readings. students prep before class, class time shifts to actual discussion. 6. deep research — NotebookLM searches the web autonomously, finds gaps in your sources, builds a cited bibliography. one researcher cut lit review time by 70%. 7. google classroom integration — share notebooks as view-only. students query, generate their own study aids, all from the same materials. what took a summer now takes a week. what took a week takes an afternoon. twitter.com/the_smart_ape/stat...
UPLOAD
0%
The Smart Ape 🔥
03-21
Thread
#Thread#
been digging into @mezonetwork before TGE and looks interesting. ~4% yield on BTC directly from protocol revenue. $80M already in the vaults and the token isn't even out yet. the part most interesting part is the fee flywheel: institutions deposit BTC → mint MUSD → generate protocol fees → veMEZO holders vote on where those fees go. more institutions in = more fees = more value for lockers. and institutions who don't want to hold $ MEZO can rent your veMEZO position, secondary yield on top of the base. $28.5M from Pantera, Multicoin, GSR. if you lock MEZO into veMEZO early you basically get first dibs on where yield gets directed. that voting power dilutes fast once everyone else come.
BTC
0.67%
Loading..