3 ms·
This whole thing makes me wonder if AI could detect "anomalies" like the human who found the actual hack did. Observe a system (lots of systems) and use that da
by EasyMark 3y ago
This whole thing makes me wonder if AI could detect "anomalies" like the human who found the actual hack did. Observe a system (lots of systems) and use that data to spot anomylous behavior from "new" versions of packages being added, throw up a red flag that is like "this really doesn't act like it did in the past because parameter(s) are unusual given previous versions"
- pixelfarmer 3y agoJust needs a fine grained enough set of parameters that are observed on a system, aka "monitoring", and flag anything that sticks out. No need for "AI" unless the parameters that are monitored have more of a (complex) pattern to them than some value +/- deviation or similar things. "AI" is, in an abstract way, a pattern matcher and replacer, but also a big cannon that is often not needed and simpler heuristics work faster and better (and are usually deterministic).
- lpapez 3y agoAnomaly, fraud, outlier detection... AI has been trying to solve that for several decades and in my experience in production these systems usually end up being a set of expert written rules saying "if X is greater than Y, notify someone to check".
- mjburgess 3y agoThe issue is that fraud is anti-inductive: when it's discovered, it changes. Statistical AI systems are just naive induction, and so largely useless.