Newmark's Door

Gary N. Smith, Fletcher Jones Professor of Economics, Pomona College, addresses something I've wondered about but haven't yet taken the time to try to figure out: how do these machine learning algorithms establish causality?

Artificial intelligence (AI) algorithms are terrific at discovering statistical correlations but terrible at distinguishing between correlation and causation. A computer algorithm might find a correlation between how often a person has been in an automobile accident and the words they post on Facebook, being a good software engineer and visiting certain websites, and making loan payments on time and keeping one’s phone fully charged. However, computer algorithms do not know what any of these things are and consequently have no way of determining whether these are causal relationships (and therefore useful predictors) or fleeting coincidences (that are useless predictors).

And here he proposes the "Smith Test" to determine if computer algorithms can make reliable recommendations.

Do tell.

Also math-free but I kinda think you need the math.

But there's this: ". . . no one on Earth fully understands the inner workings of LLM."

Chalk up one good use for AI.

May be of interest to folks who want to learn programming. (Seven-minute video.)

Related: "How I would learn to code in 2023 (If I could start over)".

I take the causality story here with several tons of salt. But if it is causal, I, for one, would be happy about it.

Probably a good list to save. I've only used three of these sites--Down for Everyone or Just Me, Ninite, and ManualsLib--but they were all quite useful, so maybe some of the others will be, someday, too.

How long before hellacious competition trims these figures a bit?

I smell a graduate paper if not a Ph.D. thesis or two: what factors account for success or failure when companies rename themselves?

It's rare, I think, but it does happen: Google let this person down.