@beefnoode: Last night I had a coffee chat with @Charles77xixi, and we really enjoyed the conversation. Teacher Charles is relatively young, but has already achieved great results in the AI field. He graduated from a top school and entered a top-tier big company, and now works at a star AI startup. In terms of technology, even though I…
Summary
The blogger shares insights from a coffee chat with AI practitioner Charles, discussing the choice between big companies and AI startups, entrepreneurial paths, differences between B2B and B2C markets, and the importance of networking and resources.
View Cached Full Text
Cached at: 06/25/26, 01:19 PM
Last night I had a coffee chat with @Charles77xixi, and it was really enjoyable.
Charles is still quite young, but has already achieved significant results in the AI field. He graduated from a top university and joined a leading big tech company right out of school. Now he works at a star AI startup. On the technical side, even though I don’t understand the details, I can tell he’s incredibly impressive.
We talked a lot about working vs. starting a business in the AI industry. A few key takeaways worth sharing:
-
The bar for landing a job at big tech companies keeps rising, but the growth potential and flexibility of those roles are actually decreasing. It’s worth paying more attention to AI startups that have secured funding.
-
There are two paths for AI entrepreneurship: One is the “grow fast, cash is king” approach. The other is the “tell a story, raise capital” approach. Neither is inherently better than the other, but the resources and capabilities required are completely different. You need to play to your strengths and leverage your unique advantages.
-
On choosing a direction for AI startups: No matter what, go for directions where “the stronger the base model, the stronger you become.” Avoid building application-layer products that can easily be undermined by the underlying model’s capabilities.
-
Looking at current market demand, using AI to improve efficiency in specific business scenarios remains the clearest direction — whether through products or services.
-
In China, doing B2B is much easier to survive than B2C, because Chinese C-end users’ willingness to pay is far lower than overseas. However, B2C products have more room for imagination and storytelling.
-
Although USD funds have largely pulled out, domestic RMB VCs are maturing. The funding window for young AI founders born after 2000 will open further.
-
Every entrepreneurial venture is actually an accumulation of your network and resources. During the startup process, you must proactively connect with external resources — investors, partners, and co-founders. Truly valuable things should be retained. Don’t keep your eyes fixed on just one project.
-
For VCs, a founder with multiple entrepreneurial experiences who has previously earned investor trust is far more attractive than a complete newcomer.
-
As powerful as AI is, ultimately it’s about people — the relationships, connections, and trust between them, because that directly determines resource allocation and direction.
Let me once again recommend @Charles77xixi — truly remarkable! He’ll most likely move toward AI entrepreneurship in the future. It’s no exaggeration to say he could very well be the co-founder of the next multi-million-dollar AI startup.
I’ve finally returned to Shenzhen and am back to meeting people in person — socializing offline like crazy! If you’re also working on an AI startup, feel free to grab a coffee chat with me. Let’s see if we can create something together.
Similar Articles
@leslieloser_: Had the privilege of meeting @Zhm20220917, the best at AI transformation in Jiangsu, Zhejiang, and Shanghai, for a few hours. Became even more certain about the following --In the AI era, those closer to production who understand the industry will reap huge startup dividends; understanding AI and boundaries is 20%, understanding production and industry is 80% --Small teams refuse inno…
The article shares insights on entrepreneurial dividends in the AI era, emphasizing that understanding industry and production is more critical than mastering AI technology. Companies prioritize actual problem-solving capabilities over the models themselves.
@10xmylife: https://x.com/10xmylife/status/2078690248645910965
Independent developer Haiming shared his experience and thoughts on running a one-person company (OPC) in the AI era, including using AI tools such as Cursor and ChatGPT to reduce production costs and improve efficiency, while emphasizing that market risks still exist, and providing an execution methodology.
@blackanger: This is the transcript of my sharing at the Data & AI Meetup organized by Databend at PingCAP office on Saturday afternoon (June 6, 2026): How I evolved from an old-school programmer to the director of an AI software factory.
The article documents a sharing at the Data & AI Meetup about evolving from traditional programming to an AI software factory, and mentions PingCAP's enterprise agent collaboration product LOOP.
@MaxForAI: I laughed when I saw this post claiming AI can't replace professionally trained people. A bunch of people are nitpicking over the fact that big influencer Kazike doesn't understand CDN details. Then they smugly say, "See? They still need us, the professionally trained ones." I don't get what they're so smug about… Hu Yanbin's Vibe coding app had some bugs recently, and a bunch of people were cheering about that too. Let me remind these friends: the reason "non-professionals" make many rookie mistakes now is essentially because AI isn't strong enough yet. Tomorrow, the day after, these bugs will become fewer and fewer. Clinging to "today's knowledge gap" as a moat is basically betting that AI won't improve. The real point here isn't whether "professionals have an advantage"—it's that for the first time, AI allows a person to cross into a completely unfamiliar field at low cost, to understand, build, and produce. This is a paradigm shift. So the question should be flipped: you people who are already experts in your own fields—why aren't you using AI or software engineering to conquer other domains, instead of sitting in your moat feeling superior by pointing out others' bugs? Don't resist evolution.
Discusses how AI development enables non-professionals to produce across fields, and that professionals should not cling to knowledge gaps but embrace AI tools.
@ba_niu80557: Let's talk some hardcore practical knowledge while I have time this morning. What actually happens between signing a contract for an AI project and it finally running in production? I'll lay out the entire playbook. People in this field can copy it directly, and those not in it can still understand why 95% of enterprise AI pilots end up dead. First, let me say something counterintuitive to the point you might not believe...
This article discusses common reasons for the failure of enterprise AI projects from proof-of-concept to production deployment, highlighting key practices such as MLOps, early inspection of real data, and clear human-machine boundaries. It argues that project failures are often not due to model issues but due to neglect of the engineering implementation phase.