It is now the morning of January 14, 2026.
Another day of waking up super early—I was awake a little after 5 AM. I fell asleep very early last night, around 9 PM. It might be because of the intense work over the past few days; my body needed rest. Although my sleep schedule is extremely irregular, I almost always get 8–9 hours of sleep.
After waking up early, I handled some emails and messages, then started writing my journal.
On Influence
Yesterday, I worked on summarizing DeepSeek's Engram paper and briefly posted a news commentary on my social feed, which sparked some discussion. It seems many people are quite interested in large language models. I didn't really produce any new insights myself; I just summarized the paper's content. I added some of my own perspectives based on my understanding.
This is essentially the benefit of following trends—as long as you can keep up with hot topics in a timely manner and produce valuable content, you can attract attention. If you gradually build up your influence, you'll have more opportunities in the future.
Our NTNL team has worked on many excellent projects in the past, but because team members tend to be low-key and we lacked promotion, many projects didn't receive the attention they deserved. In the future, we will focus more on promotion and outreach to let more people know about our projects and achievements.
I hope there will be more opportunities in the future to share our ideas and experiences with everyone.
Influence is a long-term accumulation process that requires consistently producing valuable content and actively participating in community interactions.
Ultimately, it manifests in a person's personal brand, helping them gain more opportunities and resources in the future.
This is an independent multiplier for the scale of results we can ultimately achieve, and it plays a very important role.
Simply put, the return on doing something = (pricing - cost) × influence. If influence remains only a single-digit number, no matter how high the pricing is, the final return won't be very large.
In quantitative trading, the factor of influence is often overlooked—people think that as long as the strategy is good, they can make money. Because you only need to pitch to a few major clients to secure large funds. But when working on a project, especially one aimed at the general public, influence becomes very important. The greater the influence, the more funds and users you can attract, and the higher the final return.
However, it all leads to the same destination—people with greater influence often have better resources and opportunities to work on better projects, creating a virtuous cycle.
On LegionMind
Yesterday, C1 improved the plan for the LegionMind project. LegionMind has decided to integrate GitHub Project Kanban functionality as an Agent project management tool.
This interaction is clearly inspired by vibe-kanban, but vibe-kanban built its own kanban system, whereas LegionMind directly leverages GitHub's built-in Project feature.
The advantages of this approach are:
- GitHub Project naturally integrates with production assets (code repositories). All tasks and projects can be directly linked to code, making management and tracking easier.
- Users can seamlessly use GitHub Project's kanban UI without needing to learn a new tool. Additionally, it unlocks mobile access, which is something vibe-kanban currently lacks.
- Stronger team functionality and a more seamless experience.
The trade-offs are some customization features—for example, the style and layout of the kanban board may not be as flexible as a self-built system. It also means giving up on users who don't use GitHub.
I think these are acceptable trade-offs in the early stages of a project. First, find the best workflow for yourself, break through with a focused approach, and then consider adapting to more scenarios and users. To be fair, vibe-kanban could also achieve the advantages mentioned above, but it would require additional development effort. Investing in that wouldn't be worth it.
If you start off unwilling to compromise on anything and try to make something universally perfect in every way, the result is often falling short on both ends and accomplishing nothing well.
Lately, I've realized many things—I feel that when doing something, you need to make trade-offs and can't have everything. Sometimes, you have to give up to gain.
