I spent twenty years becoming good at the wrong game
savvynormie.com
August 20, 2026
4 min read
45/100
Summary
A Russian academic who earned Candidate of Sciences and Doctor of Sciences degrees left their previous research path after concluding that its methods could not clearly explain or test its claims. The writer defended a second dissertation in February 2024, then decided by June to leave the academic system, their research field and Russia. They said their work often began with broad claims about society and used flexible theoretical frameworks to find supporting passages, leaving few clear conditions that could disprove the conclusions. In fall 2024, at age 41, the writer enrolled in computational linguistics and natural-language-processing coursework despite having no coding experience or statistical background. They described the new work as more accountable to evidence because code can fail, parsers can disagree and data can contradict an intended interpretation. The writer is now evaluating work by transferable skills, finished artifacts, usefulness, readers, opportunities and freedom rather than academic credentials alone. The writer burned the bound copy of the second dissertation and deleted most research notes from that period. They and their fiancée have been disposing of their belongings and booked one-way flight tickets scheduled to depart within weeks, without knowing where they will be six months later.
Key Takeaways
- Russia’s Candidate of Sciences and Doctor of Sciences are successive research degrees, with Doctor of Sciences serving as the higher doctorate; the writer earned both.
- The writer said publication in Russian journals offered a safer and more predictable path than seeking review at more selective international journals, where rejection was more likely.
- The writer began studying computational linguistics and natural language processing in fall 2024 at age 41, starting without coding experience.
- The writer said computational work provides clearer constraints because software failures, conflicting parser outputs and data can challenge a proposed conclusion.
- The writer and their fiancée booked one-way tickets to leave within weeks after reducing their possessions.
What the discussion said
The AI-related part of the thread fixated less on the author’s career detour than on whether the essay sounded machine-written. Several readers saw the polished but vague, highly abstract style as a recognizable Claude-like failure mode: it delivers neat philosophical framing while withholding the concrete education, research, and personal turning points that would make the account credible. Their complaint was not simply that AI prose is imperfect; it was that generic language can flatten a lived story into interchangeable self-help rhetoric, making readers suspect a generated text even when the underlying experience may be real. There was little effort to establish authorship conclusively. One participant suggested that the unusual phrasing could reflect translation or a non-native English inflection rather than model output. Still, the dominant AI-adjacent reaction was that readers wanted provenance and specificity: the actual degree, dissertation topic, and episodes of disillusionment, rather than a smooth summary that seemed optimized for thematic resonance. The wider debate about academia, social science, and career value was largely unrelated to AI and did not add a substantive assessment of models or AI research.
Where opinion split
The central dispute was whether the essay’s abstract, polished voice indicates AI generation. Skeptics argued that Claude-style prose substitutes broad, interchangeable reflections for verifiable personal detail; the counterpoint was that translation or the author’s linguistic background could produce the same effect, so style alone is weak evidence.
Community Sentiment
Positives
- Readers’ insistence on concrete personal evidence reflects a useful standard for AI-era publishing: provenance and specific experience matter more than elegantly generalized prose.
- One commenter cautioned that non-native phrasing or machine translation can resemble generated text, pushing back against treating stylistic suspicion as proof of AI authorship.
Concerns
- The essay’s abstraction was repeatedly read as Claude-like output, with generic thematic language obscuring the human story readers expected to find.
- Commenters argued that AI-flavored prose can turn a potentially compelling account into bland, interchangeable reflection, eroding trust even when the claims may be genuine.
