Companies are using LLMs to solve classification problems better handled by classical ML
FrankFelixAI · x · 2026-09-17
mstockton shares raw thoughts on the "Jev model": lots of business problems are really classification or regression problems, yet many companies don't realize classical ML can solve them and rely on people and process instead.
With LLMs going mainstream, more companies now think "maybe AI can do that!" and throw LLMs at classification/regression tasks. This automates some manual work, but LLMs are often the wrong tool and may underperform classical ML. The popularity of "AI" (now synonymous with LLMs for most people) pushes companies to default to LLMs for everything.
More from Companies & People
- Cloudflare engineer marvels that Swagger/OpenAPI inventor Zeke is his colleague — threepointone · 2026-09-17
- Max Planck School of Biomedical AI opens first PhD cohort, apply by Dec 1 — kmborgwardt · 2026-09-17
- OpenAI launches inaugural Safety Fellows cohort on alignment and interpretability — tomekkorbak · 2026-09-17
- Two kinds of platform: confusing app platforms with infra gets worse in the agent era — matt_slotnick · 2026-09-17
- David Sacks tells Amodei and Altman to slow down voluntarily, not via regulation — mikeflache · 2026-09-17
- Breakout List names 92 startups to join, 20 with 25 or fewer employees — BenBlaiszik · 2026-09-17