Advice for lost computer science students

Overview

The video argues that computer science students are often set up to struggle because “traditional” signals—degree, coding ability, and LeetCode-style interviews—no longer guarantee career traction, especially in an AI-driven job market. The creator claims their own path went from being lost and rejected to landing an Amazon internship and a $220,000 big-tech job by adopting three “brutally honest” lessons.

---

1) The “computer science degree trap” — degrees don’t get you hired; referrals do

The creator describes applying to 100+ companies with strong coursework and projects (machine learning, data structures/algorithms, learning multiple languages) but receiving no responses. Their conclusion: the issue wasn’t ability—it was that many companies don’t value the degree as much as relevant signals like networking and referrals.

They use an analogy of LeBron James effectively helping his son (Bronny) get drafted via connections, concluding that “it’s who you know.”

Action recommendations:

• Reach out to 30–50 engineers per day on LinkedIn with thoughtful messages; ask about opportunities and pursue coffee chats and, eventually, referrals—without sounding transactional.
• Build visibility by posting strong projects on Twitter and LinkedIn, so recruiters/hiring managers discover your work directly rather than relying solely on applications.

---

2) Software engineering is “pregnant” — AI is spawning new roles, and workflow leverage matters

The creator claims software engineering is splitting into emerging subfields that will grow rapidly:

• AI Engineer (example: adding an intelligent chatbot layer to a recipe app). The video suggests AI engineer roles can pay around $220,000 entry-level.
• Forward-Deployed Engineer: engineers placed closer to customers to gather real problems and build technical solutions directly (framed as a “front line” evolution).

They emphasize that the real advantage comes from using AI inside the tools where work happens—not manual tab-switching and copy/paste. A Zapier demo is used to argue for automation/agents integrated via an SDK to reliably run workflows (e.g., a scheduled “morning executive brief” pulling from Calendar, Gmail, and Slack).

Suggested skills for forward-deployed engineers:

1. Strong coding ability and debugging.
2. Ability to use AI beyond prompting—e.g., orchestrating tool-using agents/sub-agents end-to-end.
3. Product thinking (customer interaction, usability, psychology), not just writing code.

---

3) “LeetCode is dying” — interviews are shifting toward more job-relevant tests

The creator argues LeetCode has become less meaningful because candidates memorize patterns and may even cheat using AI. They cite examples of companies moving away from it (e.g., GitLab, Snapchat, Zapier) and note that Meta is experimenting with new interview formats.

They propose a new trend: treating the interview like a paid micro-contract. Under this model, candidates are given part of a real codebase to build a specific feature end-to-end. If it works, they’re hired full-time.

The creator argues this better matches actual job tasks—for example:

• Building an AI chatbot layer for an AI engineer role
• Fixing a buggy codebase for a forward-deployed engineer

---

Overall takeaway

To improve career outcomes, the video recommends:

• Prioritize networking/referrals to avoid getting lost in the crowded graduate “ocean.”
• Prepare for AI-adjacent and customer-facing engineering roles, and learn to integrate AI into real workflows.
• De-emphasize LeetCode memorization in favor of job-relevant abilities as hiring shifts toward practical, feature-based evaluations.

---

Presenters or contributors

• Primary presenter/creator: (Name not provided in the subtitles)
• External collaborator referenced: Zapier (partner mentioned)
• Interview subject referenced: Satya Nadella (CEO of Microsoft)
• Public figure referenced in analogy: LeBron James (and Bronny by implication)