@FinanceYF5: Career advice in the AI era from someone who has worked at Scale, OpenAI, and Google 1/ AI can solve all problems with standard answers School exams, Leetcode, system design — these are becoming AI's home turf What's truly valuable is the ability that can't be written as a loss function

X AI KOLs Following News

Summary

A practitioner who has worked at Scale, OpenAI, and Google shares career advice for the AI era: AI is good at solving problems with standard answers (e.g., exams, Leetcode), while the truly scarce abilities are those that cannot be defined by a loss function.

Career advice in the AI era, from someone who has worked at Scale, OpenAI, and Google 1/ AI can solve all problems with standard answers School exams, Leetcode, system design — these are becoming AI's home turf What's truly valuable is the ability that can't be written as a loss function 🧭 https://t.co/rPJe1fjtqY
Original Article
View Cached Full Text

Cached at: 07/06/26, 06:01 AM

Career advice for the AI era, from someone who has worked at Scale, OpenAI, and Google

1/ AI can solve any problem with a standard answer

School exams, Leetcode, system design — these are becoming AI’s home turf.

What’s truly valuable are the abilities that can’t be written as a loss function.

2/ He turned down a higher-paying quant offer at Scale

Through Scale, he met a bunch of future founders and landed opportunities at DeepMind and OpenAI.

His conclusion: money is the least scarce resource.

【Time, relationships, reputation】 — those are the real scarce goods.

3/ His company no longer tests Leetcode in interviews

Instead, they test: if you’re thrown into an unfamiliar environment, can you quickly find a problem worth solving?

Agents can solve complex problems, but only if someone first determines which problem is worth solving.

4/ There’s only one criterion for choosing a company

Is this company working on the most ambitious version of this problem? And do they really have a chance to make it happen?

AI makes ordinary software too easy. The only real moat is extreme focus on genuinely hard problems.

5/ The last 10% accounts for 90% of the value

An Agent’s output is average. The real value comes from that extra polish, clean architecture, creative details.

In interviews, he can tell immediately who has spent that time and who hasn’t.

6/ In 2023, he turned down offers from Anthropic and Cursor

Looking back now, he feels the decision direction was right, but he didn’t gather enough information at the time.

Anthropic’s earliest product was a Slackbot that performed worse than ChatGPT.

7/ AI will not replace all knowledge workers

Because choosing which problem is worth letting AI solve — that itself is human work.

What you’re practicing now: is it the ability to find problems, or just the ability to solve them?

Read the original text:

That’s all.

If you like this topic:

  1. Follow me (@FinanceYF5)
  2. Like + repost the first post below

Someone used Fable 5 to build a customer acquisition system for renovation contractors.

They scraped houses sold locally in the past 12 months, used visual recognition to skip yards with existing shade, calculated how long the yard has been exposed using satellite data, rendered a pergola into the homeowner’s own yard photo, and printed it as a postcard to mail out.

Each order ranges from $6,500 to $18,000. What do you think of this customer acquisition approach?

Similar Articles

Andrew Ng's Summary of 4 Core AI Engineering Competencies 1/ After analyzing over 10,000 job postings and interviewing dozens of AI experts, recruiters, and headhunters, Andrew Ng identified the 4 most critical AI engineering skills: building and deploying AI applications, software engineering fundamentals, using Coding Agents, and defining what to build...

X AI KOLs Following

By analyzing over 10,000 job postings and interviewing dozens of AI experts, Andrew Ng summarized the 4 core AI engineering skills: building and deploying AI applications, software engineering fundamentals, using Coding Agents, and defining what to build.

@ai_super_niko: https://x.com/ai_super_niko/status/2070299861757616606

X AI KOLs Timeline

This article discusses whether computer professionals still need to learn technical skills in an era where AI can write code. The author argues that surface-level technologies like syntax and APIs are depreciating, but deeper capabilities such as algorithms, design architecture, and judgment become more important. The focus of learning should shift from beginner-level skills to the knowledge required of senior engineers.