rage baiting men on this app and farming interactions by pissing them off only for them to shout “onlyfans wh*re!” from the rooftops, advertising & further promoting me for free like the good little sleeper agents they are. god is good.
I decided to strictly implement my own ideas that only promote English interaction.
Many Chinese users in Chinese platform complain to English users when they use English. Their arguments is: This is a Chinese platform so please respect us by using Chinese.
OK, let us be consistent in rules: X is an English platform and everyone should use English to respect the most broad users. In particular, X should not be accessed by Chinese law. ENGLISH AND GERMAN USERS also dislike Chinese letters saturated here as I keep receiving feedbacks. I also need to respect their feelings.
The trade-off is: this account is more meaningful to let international users see what my nice life in China is and make them no more trust western propaganda. This is a real constructive thing. For that purpose, Im happy to dump the subscribers who cannot understand the complexity of international communication.
Besides, I post most of contents already in Chinese platforms. I know what I'm doing and I'm not purposely pissing anyone. What I gain from it? There are must be a reason behind it.
The hard fact is the oversatured Chinese comments begin to cause this account less attractive to international users. I dont know who is good willed and who is bad willed behind those comments.
If you have better suggestions, please share.
Former New York Governor @andrewcuomo, a board member at crypto exchange @okx, calls for Congress to pass the CLARITY Act focused on regulating cryptocurrency.
"The world is passing us by," he says.
We tried using Meta's new Muse Code agent, but it has a bug that doesn't let it sign in from a docker container.
So we did a fun experiment: Meta claims Muse Spark 1.2 was co-trained with their Muse agent harness. So we extracted instructions from their system prompt and added them to the Cline harness.
TL;DR of this special prompting:
- Trust source code over the user prompt, so read every call site and existing tests before starting the task
- Weigh edge and error cases as heavily as the happy path
- Always reproduce the bug before fixing
- Don't trust the first passing test suite, and verify suspicious looking half-baked tests
- Never stop at just editing, keep working until the change is verified complete.
We then asked this modified harness to fix a real bug from our repo, and compared the results to the original Cline agent harness.
Results:
- Used 2.7x fewer tokens (19.7M → 7.2M)
- Finished 2x faster (49min → 24min)
- Cost 2.4x less ($7.69 → $3.25)
Same Muse Spark 1.2 model, same task, only the prompting changed. Incredible how much of a performance gain Meta was able to achieve training it on these special instructions!