Most important take away

AI has lowered the bar for doing average work, which means average no longer stands out. The way to be hired in 2026 is to deliberately do the inconvenient, uncomfortable work most candidates skip: solve real problems for real people, get warm referrals through second-order connections, learn one technology deeply (not 50 shallowly), and create your own experience by shipping a real business that earns at least one paying customer.

Summary

Steve Huynh, a former Amazon principal engineer of nearly 20 years, argues that AI has not made tech hiring harder; it has exposed how many candidates take the convenient path. Because tutorials, resume polish, LeetCode answers, and portfolio projects are all trivially generated now, the differentiator has shifted from “can you build” to “whose problem can you solve.”

Actionable career advice from the episode:

• Stop building generic showcase projects (chatbots, to-do apps, Titanic predictors). Instead, ask three people in your network what the most annoying repetitive part of their job is, then build a solution. Replace resume bullets like “built a chatbot in Python/Flask” with measurable impact such as “reduced manual data entry by six hours per week for a real estate agency.”
• Networking through cold LinkedIn requests is the comfortable, low-yield version. 60-70% of jobs are filled via networking and referrals, the one channel AI cannot replicate. Search a target company on LinkedIn, ignore first-order connections, and look for second-order connections (mutual contacts). Ask your mutual for an intro, then in the conversation ask for advice (team focus, what they look for, tech stack, prep tips), not the referral. Ask for the referral at the end. If ghosted after applying, your referrer can ping the recruiter internally.
• Stop going wide across 50 technologies. Pick a tool already in your stack, read the latest release notes (not tutorials), find a new feature most people have not learned yet, build something with it, and then publish a blog post or video teaching it. This builds real depth, validates your understanding, and makes you discoverable to recruiters and hiring managers.
• Solve the experience catch-22 by manufacturing your own. Form an LLC (LegalZoom or similar), get an EIN, open a business bank account, set up Stripe, then ship a small product and get one stranger to pay one dollar. That single transaction reframes you from “applicant with side projects” to “founder with paying customers,” and teaches you marketing, pricing, support, and shipping with real stakes.

The recurring pattern: the work that wins is specifically the work that is inconvenient, ambiguous, and unguided. AI removed the friction from everything that has a tutorial, so the remaining moat is doing what has no tutorial.

Tech patterns mentioned: AI-assisted coding has a quality gap between one-shot generated code and a polished product a stranger will pay for; that gap is the opportunity. ATS filters and AI screening can be bypassed by warm human referrals. Reading release notes (vs. courses) is positioned as the fastest path to genuine expertise because new features have no Udemy course yet.

Chapter Summaries

1. The new game - AI made average work easy, so doing average things is no longer impressive. Outliers win by doing the uncomfortable work others avoid.
2. Problem 1: Projects nobody cares about - Tutorial-style portfolio projects are noise. Ask people in your network about their painful repetitive tasks and build a real solution with measurable impact.
3. Problem 2: Fake networking - Cold LinkedIn requests are comfortable but low-yield. Use second-order connections, request advice first, then a referral, and follow up through your referrer if ghosted.
4. Problem 3: Going wide instead of deep - Listing 50 technologies makes you invisible. Pick one tool, read its release notes, master a new advanced feature, build with it, and teach it publicly to build depth and a personal brand.
5. Problem 4: Waiting for experience - Form an LLC, get an EIN, open a business bank account, set up Stripe, ship a product, and earn one paying customer. Walk into interviews as a founder with traction, not a candidate with tutorial projects.
6. Closing - The bar to be an outlier is on the floor because being impressive requires inconvenient work most people refuse to do.