Self-taught engineers often outperform (2024)

Passion and motivation vs. learning path

• Many argue the real differentiator isn’t “self-taught vs degree” but passion, curiosity, and willingness to keep learning.
• People who tinker on side projects or pursue hard topics on their own often retain concepts better and map theory to real problems.
• Several note that both strong CS grads and strong self‑taught devs share this trait; weak performers exist in both groups.

Formal education: benefits and gaps

• University is praised for: forcing students out of their comfort zone, exposing them to fundamentals (DSA, OS, networking, math), and giving shared vocabulary.
• It can fill “boring detail” gaps that self-taught devs often miss initially (e.g., complexity, data structures, concurrency).
• But many CS programs are criticized as theoretical, outdated, or shallow on real-world engineering (large systems, tooling, performance, debugging).

Self‑taught path: strengths, weaknesses, survivorship

• Self-taught devs are seen as naturally filtered: only those who can actually deliver tend to break into the industry.
• Strengths mentioned: persistence under uncertainty, comfort with learning new stacks, practical problem-solving, high output, and broad, idiosyncratic knowledge.
• Weaknesses: missing fundamentals, uneven skill “spikes,” reinventing wheels, difficulty knowing what they don’t know, and stronger impostor syndrome.
• Several note that after ~5–10 years of experience, differences in initial path often blur.

Hiring, credentials, and bias

• Degrees are described as a blunt but convenient hiring filter and a proxy for baseline competence and socialization.
• Some managers prefer experienced self-taught candidates over fresh grads, others the reverse; many emphasize a mix of backgrounds on teams.
• Cost and access to university (especially in the US) are raised as major class filters, separate from ability.

Theory vs. practice and domain differences

• Multiple comments stress you “need both”: theory to recognize and frame problems, practice to ship and maintain real systems.
• Distinction drawn between software and licensed fields (civil, mechanical, etc.) where formal credentials and standards are non‑optional.

Critiques of the article and terminology

• Several point out survivorship bias and lack of data; the headline is seen as over-claiming.
• The examples (Linus Torvalds, Margaret Hamilton) are criticized as actually highly educated.
• “Self-taught” is often reframed as “informally educated” rather than literally learning in a vacuum.