LLMs 4
- Explanatory text and jargon generated by machines for readers with different levels of understanding makes communication worse.
- Local models are not yet reliable enough for production software development.
Vicki Boykis
Machine learning engineer working on recommender systems, personalization and information retrieval; author of the 'What are embeddings?' text.
Vicki Boykis did not write this page.
We collected these quotes from things they published elsewhere, and every quote links to where it was said. They have no account here and have not endorsed this site. Quotes are word for word; the short line under each one is our own restatement, not their wording. Their own site. Is this you? Claim it or ask us to remove it. Or tell us what is wrong here.
What Vicki Boykis says most about
software quality 3
- A feature added to a codebase is extremely difficult to remove once other code is built on top of it.
- As AI makes it easier to generate code, teams must delete code more aggressively or be overwhelmed by accumulated complexity.
code generation 2
- AI code generation's default reward loop works like a slot machine and undermines skill retention.
- As AI makes it easier to generate code, teams must delete code more aggressively or be overwhelmed by accumulated complexity.
documentation 2
- Explanatory text and jargon generated by machines for readers with different levels of understanding makes communication worse.
- AI-generated technical documentation is optimized for completing a sentence rather than for being understood by a human reader.
simplicity 2
- An inability to explain something simply often reflects not yet understanding the problem well enough to compress it.
- Deleting code goes unrewarded in software careers, which is part of why teams keep shipping new features instead of removing old ones.