Software engineers can no longer neglect their soft skills

AI, learning, and cheating

• Mentors report a split: strong CS students use LLMs to deepen understanding; weak ones use them as a crutch and can’t explain or recreate “their” work.
• Code cheating is easier to detect via oral questioning; essay cheating is harder, though some suggest viva-style defense of essays could work too.
• Others note that many honest writers can only give post‑hoc reasons for structure and focus, so “explain your choices” isn’t a foolproof test.

Hard vs soft skills in the AI era

• One camp argues “real” engineering skills will matter more: math, low‑level understanding (C, assembly, Linux, devops), performance, simplicity, resource efficiency.
• Another camp emphasizes higher‑level abstraction, systems thinking, product/UX design, domain modeling, and measuring impact as the differentiators.
• Several commenters say AI excels at “autistic” code‑generation, so remaining human value will skew toward communication, negotiation, framing problems, and turning fuzzy requirements into robust systems.

Soft skills: always needed or newly critical?

• Many insist soft/professional skills were always necessary except for rare “generational talent” outliers; weak communication has long been a career tarpit.
• Others say there were viable niches for “ticket takers” who just implemented scoped Jira tasks; AI plus cost pressure may now erase these roles.
• There’s debate over “brilliant jerks”: some teams tolerate them for difficult problems; others report they slow projects and poison morale.

AI, productivity, and work expectations

• A subset claim 2×–50× productivity gains with tools like Claude Code and see AI as an “exoskeleton” for serious engineers.
• Others report small gains (+5%) or find current tools useless beyond trivial boilerplate, and warn that LLM‑generated code is harder to maintain.
• Some push back: if you do 2× the work, expectations rise without 2× pay; AI doesn’t fix bad requirements, shifting goals, or politics.

Organizational dynamics and “soft-skill fortresses”

• Several warn against organizations where soft skills/politics outweigh execution, seeing that as a sign of capture by weak leaders or “bullshitters.”
• Others counter that effective execution requires both: aligning people, resolving misunderstandings, and maintaining cohesion are core engineering work, not fluff.
• Renaming “soft skills” to “professional” or “durable” skills is suggested to reflect their difficulty and importance.