GateUser-d796ae8d

In the vibe coding era, if anyone tells me "I have an idea," I'll punch it flying.
If you have the effort to talk about it, you could've already vibed out an MVP.

I feel like Xiaohongshu's feed algorithm is really weird. When I post legitimate content, it absolutely won't get any traffic (OPT RFE strategies, rental information, etc.). So what content actually gets pushed to me? A squirrel stealing my stuff, using borderline copy to describe life phenomena, complaining about expensive parking in Seattle, my handsome photos (non-borderline) 🤡

Recently I've been liking to delete friends & block some content on my timeline, which includes but is not limited to:
1. I set up xxx with xxx in 10mins, … I'm going all in on AI!
2. It's not…, it's not…, it's…

Recently, whether it's replying to emails, reviewing manuscripts, or writing commit messages, I usually hand-type the important paragraphs myself, setting aside the boilerplate. As for trivial typos and formatting issues, I simply let them be.
AI has made precision cheap, while the human touch of traditional craftsmanship grows ever more scarce.
—— This article was polished and published with Claude Sonnet 4.6

Everyone is talking about AI, so I'll make a prophecy.
Tools like OpenClawd(ex Clawdbot) are essentially unstable solutions built by reverse-engineering APIs and CLI. In the long run, more and more apps will provide standardized interfaces for AI to call (MCP). But this also raises another issue: entrusting user interaction to AI will inevitably weaken certain types of applications(especially toC apps)'s control over user attention, behavioral data, and so on. Therefore, in the short term, these apps will tend to avoid or actively create obstacles.
Regarding the development of agents, ther

Top 5 Favorite Video Account Heroes of 2025: Oxford Zhu Zhu Harvard Yinuo Harvard Cindykelly Want to Stay Healthy Hu Ke'er Choose Your Hero

After some thought, over the past year I have met CoinYuan, TennisYuan, CESYuan, ClimbingYuan, and InterviewYuan, but I haven't met AcademicYuan at academic conferences. Meanwhile, academic conference venues are filled with highly educated, high-income hot nerds, indicating that AcademicYuan is a blue ocean market. But where is the barrier to entry? Is it the inability to afford the conference fees?

At NeurIPS, no one had a more impressive presence than Jump Trading.

Before bed, I calculated an account and found that I had failed again.
Study abroad expenses 1 million.
The BTC price at the time of deciding to go abroad was 25,000🔪. If I had gone all in on BTC back then, it would be worth 4,000,000 now.
And I work like a cow or horse to earn compensation for mental distress. Assuming no promotion, after deducting daily expenses, I can save 50 in a year. I would need to sell for another 8 years to have 400. 🤡
Conclusion: I didn't outperform BTC

By the end of 2025, looking back at the small goals I set this year, I feel like I have accomplished nothing:
1. Contributed code to 3 well-known projects, 1 was not completed, and 1 was not merged.
2. This year, a paper was accepted, and I wanted to do an extended work for ICDE, but I missed the deadline. I estimate that IJCAI is also unlikely.
3. I trade cryptocurrencies and want to write a personal app based on previous successful experiences, with the development progress at about 20%.
4. I wanted to invest in osdi/mlsys with my classmates, but I also missed the opportunity.
5. I want to h

The fall is so bad, I want to buy the dip on some alts.
After browsing through CMC, I found only two useful projects: pendle, jito.
So this round has been all about playing with the poop.