Jack Guo

山重水复疑无路

I’m working to make that insight more accessible without having to get a job first

Had my first ever AI interview with @hf0, much better than I expected! AI even stutters a little, very realistic. Pretty fun

Couldn't agree more. The biggest moat in AI will be organizational innovations. When each individual can become 10x productive, the bottleneck shifts to human harness: how you attract, cultivate, and organize exceptional talent -- that's the next hard problem. I think the most successful companies of tomorrow will look very different from today's. They will internalize world-class training, instead of competing for the same fixed pool of candidate. They will internalize incubation and venture formation, instead of loosing ambitious talent to starting their own things. In other words, the next great company may look like an enterprise, school, and incubator combined. Pure enterprises, pure schools, and pure incubators will struggle to compete with that. ResDev is our first step toward this, starting with training.

The AI implementation war is speeding up. Anthropic’s reported $1.5B joint venture with Blackstone, Goldman Sachs, Hellman & Friedman, and OpenAI’s reported $4B+ Deployment Company backed by investors including TPG, Bain Capital, and Brookfield are a clear signal: the next phase of AI won’t be won by model access alone. It will be won in implementation. The bottleneck is not whether AI can help companies. It’s whether someone can get inside messy operating reality, understand the workflows, connect the systems, earn trust from the team, and turn AI into measurable business outcomes. That work is not glamorous. It’s not a demo. It’s implementation. Large enterprises will have no shortage of firms chasing them. The more interesting question is what happens below the Fortune 500: mid-market companies with real operating pain, real data, and real ROI potential, but without unlimited transformation budgets. The gap: how to make AI implementation economically viable for companies that are too complex for generic SaaS, but too cost-sensitive for heavyweight consulting. The opportunity is not “AI strategy” or another dashboard. It’s practical deployment against metrics that matter. axios.com/2026/05/04/ope…

Mark Cuban on the next big job students should focus on: Most companies don’t know how to implement AI, especially small businesses. “Companies don’t understand how to implement AI right now to get a competitive advantage… learn to customize a model, walk into a company, show the benefits. That is every single job that’s going to be available for kids coming out of school.” Don’t just study AI. Make it work inside a business. From our August 2025 interview with @mcuban.

A fantastic article on what Forward Deployed Engineer is and isn't, and whether you should hire one. A summary: What an FDE is A production engineer embedded inside a customer to ship the "last mile": the integration and edge cases that decide whether a deal sticks. Invented at Palantir. Listings up 800% in 2025. What an FDE is NOT — Sales engineer. SEs sell the product. FDEs build inside the customer. — Implementation consultant. ICs deploy a known product to spec. FDEs discover what to build. — Customer success manager. CSMs maintain accounts. FDEs ship production code. — Solutions architect. SAs design within product constraints. FDEs invent past them. When you actually need one — You sell upmarket. Fortune-500 ACVs, not product-led growth / freemium. — You're NOT opinionated about how customers should use your product. With a strong product vision, you don't need an FDE, you need a PM and faster customer calls. — Your ICP is heterogeneous. Different industries, different problems. A tight single-wedge ICP is a bad fit. What makes a good FDE — No playbook. Pedigree is often anti-signal: 10+ years at FAANG usually means a senior pattern-matcher who gets stuck on novel problems. — Grit. A "willingness to eat pain." — Staff-engineer-level technical bar. They could be a senior IC at any top company. — Compulsive builder. Can't help but ship. — Genuinely curious about how businesses work. review.firstround.com/so-you-want-to…

@jove 我觉得FDE需要的技能正是当下AI时代最重要的技能：judgement, taste, agency。得像PM那样思考，像consultant那样交流，像engineer那样工作，很有意思

这个周末写了一篇略带中二的长文：《我在 AI 异世界重生为 #FDE ，来自北美 #AIAgent 公司的⼀线实录》。 FDE 到底干嘛、为什么越来越多 AI 公司在招？一句话：把 AI 产品量体裁衣，送到“最后一公里”，并对结果负责。欢迎拍砖👇 superlinear.academy/c/resources/ai…

Everyone wants forward-deployed engineers now. Fewer engineers actually want to do the work. That gap matters because the hard part of AI adoption is no longer model capability. It’s the last mile: understanding workflows, translating ambiguity into software, earning trust, handling edge cases, and shipping inside real constraints. These are not “soft skills.” They are core engineering skills in the AI era. They’re also founder skills. And they’re hard to train in a classroom. They develop through real deployment: real users, real ambiguity, real stakes. That’s what we’re building at The Residency: engineers trained through deployed work, while giving companies access to talent who may not have otherwise considered the FDE path. linkedin.com/news/story/for…

Here’s the human stack - the most important skills in AI ear- do you agree? 1. Scoping problems under ambiguity. 2. Customer empathy and jobs‑to‑be‑done thinking. 3. Product judgment, deciding what not to build. 4. System integration and architecture judgment. 5. Rapid prototyping with AI tools in unfamiliar codebases. 6. Stakeholder navigation across competing priorities. 7. Edge case paranoia. 8. Evaluation and measurement for LLM systems. 9. Communication and narrative that make change land.

In the AI era, I think we need a lot more “human stack engineers” — engineers with irreplaceable human skills. Someone who thinks like a PM, communicates like a consultant, and builds like an engineer

Maybe two sided payment model could work on a small scale, but I realized I just didn’t want to work with people who were not as fully devoted as I was - externs or employees alike

Hi Twitter! I'm back

@yijiefeng Haha very much so

last minute costume idea, is this too niche?

@ZhenghuaYang @SerenityForge @_Satchely @NRGoodnight Let’s go Serenity Forge!

BOOM. #1 Trending Free on Steam front page, only 8 hours since our demo went live. We are absolutely popping off. 🙌 Thank you all SO SO much for all your support. I'm so glad Fractured Blooms is resonating so deeply with so many people🌺 store.steampowered.com/app/4029780/Fr…

@yijiefeng Haha why muted Stanford

@yijiefeng Haha, is that me?

new VC strat just dropped: charisma arbitrage

Our friends at RealXP Lab have a 12-week Game Porting Externship now taking applicants! 🎯 Solid game programming fundamentals required. 📅 09/29 - 12/19. Part-time. Fully remote. 👉 Learn more at: realxplab.com