Things you thought you didn’t need to care about that have a big impact on your job
summary
Describes concrete surprises: a naive Bayes spam example gives ~60% “spamicity” from a single token; a 99%‑accurate static analyzer still yields only ~50% probability of clean C code on a buggy corpus; hardware trends (Intel machines with 128 cores) guarantee concurrency. Recommends Bayesian thinking for iterative estimates, highlights Quarkus reactive core + virtual threads (Loom) as a way to hide concurrency, endorses AOT via GraalVM for fast startup, and suggests using the ML Chem key pair generator for post‑quantum safety.
AI-written summary · about
chapters (19)
- 0:00Invisible forces shaping your job
- 1:04Computers are hard to work with
- 3:55Systems thinking and uncertainty
- 5:19Statistics simplifies complex problems
- 7:09Introduction to Bayesian statistics
- 10:42Spam detection with Bayesian filtering
- 16:26Bayesian math isn't useless
- 18:16Diagnosing chickenpox vs cowpox
- 21:06Static analysis false positives
- 26:10Concurrency gives headaches
- 31:03Quarkus and concurrency
- 34:15Virtual threads past failures
- 35:39AOT compilation in Java
- 37:09History of encryption
- 41:47Quantum computing risks
- 44:13Venture capital hype
- 46:45Job market feels tough
- 50:39AI layoffs and Jevons paradox
- 55:14Key takeaways
related talks
53mThings You Thought You Didn’t Need To Care About
53mHow Not to Be Wrong in Software Engineering
45mDynamic vs Static Having a Discussion without Sounding Like a Lunatic
44mGrasping Complexity with Both Hands
36mHarnessing the Power of Nondeterminism for Software Engineering
55mPanel: Where Is SW Development Going?
01h 32mKeynote: It Depends..
32mAI Engineering Anti-patterns