Summary of "Will you be left behind if you don't use LLMs to code?"

Main ideas / concepts discussed

The video responds to a vague question: “Will you be left behind if you don’t use LLMs to code?”
- Dave says the question is ambiguous and can mean different things, so he breaks down multiple interpretations (hobby vs. job, learning vs. shipping output, etc.).
Consider the source (bias transparency)
- Dave introduces himself and explicitly states his bias: he enjoys programming and does not use AI/LLMs for coding on his channel.
- He argues viewers should always consider where an answer is coming from and what biases might shape it.
“Left behind” depends on what you mean by it
- The phrase can relate to:
  - Getting hired / competing for jobs
  - Keeping a job
  - Class performance / grades
  - Producing faster “finished product” code
  - Learning to understand concepts (learning outcomes vs. output speed)
- He emphasizes that comparison requires knowing the “goal.” Without defining the goal, “I’m behind” becomes a source of unproductive negative emotion.
Hobby case: you’re not “left behind”
- If programming is a hobby, Dave argues you can’t reasonably say you’re “behind,” because hobby success isn’t measured by speed or tool adoption.
- His analogy: someone isn’t “behind” in bodybuilding for not using steroids—because the hobby is about enjoyment and personal engagement, not matching competitors’ metrics.
Job/school case: possible technical “falling behind,” but it’s nuanced
- For jobs, he says “falling behind” may happen in certain environments—especially where performance is measured by output speed and the company heavily expects AI-assisted development.
- He criticizes common “dumb” metrics like:
  - Lines of code
  - (He implies) similar proxy metrics that ignore real impact
- He argues better measures would be:
  - What you shipped
  - Downstream effects / real project outcomes
  - The actual code and results, not “how long you were typing” or how many tokens you produced.
Hard constraint analogy: LLMs produce text/tokens faster
- If a company judges you mainly as a token/code-output emitter, then you likely can’t outpace an LLM:
  - He compares this to a forklift lifting more than a bodybuilder, or a car running faster than a human runner.
- But he suggests that whether you “fall behind” depends on whether the company truly values raw token/output speed versus real engineering outcomes.
AI’s effect on skill distribution: concern about “eroding the middle class”
- Dave claims he observed a pattern at work:
  - Good programmers use AI to become even better/faster (a “multiplier” effect).
  - Weaker programmers may generate worse code without improving fundamentals (and may outsource thinking/critical reasoning).
- He says this could create a widening gap:
  - Strong become stronger
  - Middle gets eroded
- He also warns that if you use LLMs to supplant your thinking (outsourcing critical thinking), that can lead to being “left behind” long-term—especially at junior levels (intern/junior).
Ethics/legal training material mentioned but not the main focus
- Dave notes ethical concerns exist (e.g., copyrighted material), but says they’re not his “biggest issue” for this topic.
- He points out that much training may come from open-source code and/or can be generated locally, which reduces what he frames as his own immediate ethical objections (though he says a fuller ethics video could be separate).
Mindset advice: manage “left behind” feelings
- His closing guidance is psychological rather than technical:
  - If you feel left behind, spend time understanding your feelings
  - He recommends learning to meditate / sit with thoughts / reduce negative emotion carryover
  - He argues that even if the fear is partly justified, carrying it around is harmful and not helpful for progress.
Channel content / personal projects (non-core but present)
- Dave shares a Patreon shout-out and describes a fun multi-language terminal pipeline project:
  - A bash script that forks a Rust program to read a CSV
  - Pipes output to a C program to write to the screen
- He emphasizes “no AI used” and that bugs came from writing it from scratch.

Methodology / instruction-like guidance (detailed bullets)

How to interpret the “left behind” question

Identify what “left behind” refers to by clarifying the goal:
- Job acquisition (getting hired)
- Job retention (keeping a job amid AI adoption)
- School/class grading
- Learning understanding vs faster production of committed code
- Whether “finished product” means:
  - Shipped code committed to a repo, or
  - Personal learning mastery
Determine the context:
- Is programming your hobby or your job?
- Are you comparing yourself to peers in a competition-like way (job market) or in a non-zero-sum way (class grades)?

When “falling behind” is more likely (per Dave’s framing)

If your company:
- Judges outputs using simplistic speed/proxy metrics (e.g., token output, lines of code)
- Treats programmers primarily as fast text producers
Then:
- You may be technically outpaced by LLM speed
- You might be “behind” relative to those metrics

How to avoid long-term skill erosion (implicit practice-oriented approach)

Don’t outsource thinking/critical reasoning entirely to LLMs.
Practice core programming skills directly (“put your fingers to the keyboard”):
- Dave frames this as exercising a skill like conditioning in the gym.
For juniors/interns especially:
- Dave believes heavy reliance risks not honing fundamentals, harming long-run competence.

Emotional/psychological response if you feel behind

Treat “being left behind” as a feeling to analyze:
- Spend time understanding why you feel that way
Learn coping methods:
- Meditation / sitting with thoughts
- Letting go of negative emotions so they don’t impair your ability to act

Speakers / sources featured

Dave (creator/host; “You Suck at Programming”)
No other external sources are featured in the provided subtitles (no named studies or cited references).