Blank wall after removing a decal mural.
Bonus points: find the snail, squirrel, and butterflies.
AI vs ML
Are there any common forms of AI now that aren’t machine learning?
Historically we had expert systems, formal logic, hand-authored decision trees (eg old school machine translation), domain-specific algorithms (eg A* pathfinding), and more.
Nowadays it seems like everything is some form of ML trained on real world data.
(I’m broadly lumping everything generated automatically from data into the “ML” bucket, including HMMs, Bayesian networks, stochastic modeling, etc.)
Maybe this is a good thing? The results definitely speak for themselves. I might buy that the scale of most modern problems simply overwhelms hand-authored approaches. Are we losing anything by putting all our eggs in the big-data-and-opaque-emergent-models basket, though? Are there any other forms of AI that we’re neglecting? Or am I just being old fashioned and naive?
Fascinating: distributed systems techniques now also apply to performance sensitive applications on high end multi-core computers, eg AAA game engines on modern consoles, as well as to actual distributed systems.
Using the holidays for a 2-week coding session. Writing a new transactional memory manager that’s O(1) for all operations, completely non-blocking and progress-guaranteeing, and uses atomics only for writes.
— Tim Sweeney (@TimSweeneyEpic) December 26, 2019
Notably, serializing distributed transactions algorithmically, offline.
The “trick” is to store a 10-bit transaction id inside variables and use that to resolve conflicts. Any transaction can undo any earlier one, and can be undone by any later one, at any time, with detection at read and write points.
— Tim Sweeney (@TimSweeneyEpic) December 26, 2019
I wonder what William Thackeray would think if he knew that his seminal 19th century novel was now being used to hock pop culture magazines, paper napkins, and underwear.
Reverse Turing tests
I’ve been watching FailRace‘s Survive the Hunt series recently, and it’s fascinating. Dennis and his friends have invented a fundamentally new kind of hide and seek. They start by jumping into GTA 5, an online game with a convincing simulation of a busy city. They turn off the map, names, and other indicators that show where players are. Dennis gets a head start, changes his clothes, finds a new car, and tries to blend in with the AI-controlled traffic and pedestrians going about their business. His friends team up and try to find him.
Most of them grab cars and race around the city, keeping an eye out for anything suspicious or out of place. A few find helicopters and blimps and take to the skies, tracking the action from above. They communicate in a group voice chat, coordinating their search and calling out sightings: make and model of car he’s driving, clothes he’s wearing, where he’s headed. Continue reading