Things You Thought You Didn’t Need To Care About
summary
Bayesian math is used to turn word counts into decisions—example: the Ozempic spam example shows a word can be 60% likely spam but a 99% accurate tool can still be wrong depending on base rates (C code example drops posterior to ~50%). Hardware forces software: core counts rise toward 128 cores, so concurrency matters; Quarkus reactive core hides complexity while still supporting virtual threads. AOT via GraalVM becomes attractive for cloud startup latency; Java 24 quantum-safe crypto is now available.
AI-written summary · about
chapters (17)
- 0:00Intro
- 1:05Geography 101: Norway
- 2:59How to make decisions in a complex world
- 4:05Statistics 101: Bayes' theorem
- 19:16Statistics in different industries
- 21:27Top tip
- 22:35Why does concurrency make my head hurt?
- 23:48Electrical engineering 101: Chip design
- 31:02Virtual threads
- 32:09Why did green threads flop & loom succeed?
- 33:37Meteorology 101: Cloud
- 34:56Physics 101: Quantum computing
- 38:53Business studies 101: Venture capital
- 42:00Macroeconomics 101: Interest rates
- 46:31Economics 101: Jevon's paradox
- 50:59Summary
- 52:06Outro
related talks
56mThings you thought you didn’t need to care about that have a big impact on your job
46mThe future of development: Are our jobs getting harder or easier?
53mHow Not to Be Wrong in Software Engineering
44mGrasping Complexity with Both Hands
04mHow it felt like
55mKeynote: The dangers of probably-working software
01h 32mKeynote: It Depends..
34mA Java Developer's Survival Guide for the Cloud