2025 has flown by so quickly, and the older I get, the faster time seems to pass.
This year felt a bit too uneventful—aside from a few moments that broke the routine, most of my life was as bland as water, so there wasn't much worth recording.
I finally managed to publish my first (conference) paper this year. Although it's somewhat recognized in the coding community, it's still not an SCI paper—keep pushing forward. However, the serious issue now is that there's no "me" attending the conference, which might lead to retraction. Ahhh, that would be truly disastrous. I really hope I can appeal successfully.
Also, the AI Codec direction I've been working on lately seems to have less room for improvement. This year, I've tried exploring new areas, such as optimizations related to large models, but so far there's been no progress. Some previous projects are still pending results, and I'm also under pressure from graduation requirements and the mandatory joint training program. The stress is real. Hopefully, next year brings a breakthrough.
This year, I finally had a few opportunities to travel beyond work-related trips. Besides multiple visits to Shanghai for various exhibitions, I went to Harbin at the beginning of the year, and during summer vacation, I finally made it to Japan—the place I'd always dreamed of visiting. I also visited Shenyang, Dalian, and Macao, but since those were work trips, there wasn't much chance to enjoy them.

Winter Harbin is definitely worth visiting—especially for someone from the south who rarely gets to see such scenery (though the tourist spots are incredibly crowded). Experiencing temperatures below zero or even minus twenty degrees Celsius for the first time, stepping into a world of frozen northern landscapes—it was quite an unforgettable travel experience.
My summer break didn’t start until late August, so I took advantage of the cheap flights and accommodations (a 5-day, 4-night trip for just over ¥2,000) to visit Kyoto and Futo, doing a pilgrimage to anime filming locations. Walking through places where characters from my favorite shows once lived, the line between fiction and reality blurred. That kind of experience already transcends mere sightseeing.



Drawing on my past experience with Next.js, I’ve been experimenting more this year and have gained some hands-on experience using AI to develop websites. As a result, I joined a startup led by a senior classmate, helping out with operations and development tasks.
Additionally, I built a small personal website—RD Curve AI—to help researchers in the compression field analyze experimental results, record data, and plot RD curves. If you’re interested, feel free to try it out—I’d also welcome any feedback.
In the past, I used to think many people were reluctant to adopt new technologies, stuck in their comfort zones, clinging to outdated tools and methods. But this year, AI has advanced so rapidly that I myself don’t have the energy to explore every new tool. What can n8n’s workflow automation do for me? How do newer models improve performance, and how do they help me?… While AI programming has genuinely expanded my development capabilities, other advancements seem to be subtly reshaping my life—yet I’ve become increasingly numb to them.
Last year I said "contentment brings happiness," but the rapid passage of time this year still left me slightly anxious. I hope I push harder in 2026 and keep moving forward faster.