Tom Cools
where else to find them
bluesky:tomcools.bsky.social
what they talk about
all 6 in the filters
where they keep going back
- Jfokus2 editions2022–2023
all 6 in the filters
on a stage with
- Baruch Sadogurskyonce
- Elien Callensonce
who talks about the same things
- Chris Simon18 talks
- Mario Fusco50 talks
- Andres Almiray57 talks
- Aja Hammerly31 talks
- Trisha Gee41 talks
- Dmitry Vinnik25 talks
| title | conference | date |
|---|---|---|
| LLMs can't optimize schedules, but AI can! | Voxxed Days CERN 2026 | 2026-02-10 |
Argues LLMs fail at combinatorial planning (30 employees → 150 shifts, search space ~30^150) and shows a working alternative: metaheuristics via the Java… | ||
| Braving the Change: Recalibrating to New Realities | JCON EUROPE 2025 | 2025-05-12 |
Students produce more code output when handed LLMs, and quality improves only when models are used as tutors; comprehension falls when code is copy‑pasted… | ||
| LLMs can't optimize schedules, but AI can! | Devworld Conference 2025 | 2025-02-27 |
Claim: LLMs fail at real scheduling because of context window limits and sheer combinatorial scale — a 10‑employee week can have 10^63 combinations — so… | ||
| #BuildPropulsionLab at #DevoxxUK: @jbaruch with @tcoolsit | Devoxx UK 2024 | 2024-05-08 |
Argument that staying current comes from fundamentals not chasing every new hype, illustrated by recognizing testcontainers in five minutes because Docker and… | ||
| Leaving a legacy: a guide to being kind to those who come after | J-Fall 2023 | 2023-11-09 |
Concrete steps to avoid leaving a rotten codebase: put docs in repo (markdown, versioned, reviewable, use Arc42), write short good‑first‑issue guides or… | ||
| Learning Through Tinkering | Jfokus 2023 | 2023-02-06 |
Claims that relevant IT knowledge has a half-life 2–5 years, and concrete experiments using Java/JavaFX for a small physics engine (location, velocity… | ||
| Modern Testing Patterns | Jfokus 2022 | 2022-05-02 |
Shows concrete ways to stop fake green tests by using mutation testing to reveal that 85% line coverage can hide a 20% mutation kill rate, using PIT and… | ||
| nothing matches | ||