Oxymoron - part the first

Amidst all the cuming about AI from the illiterate and innumerate, we finally get some sense from a labour economist out of MIT. Of course, in the world of econ, where the mantra is "it's all micro, and nothing but micro", the very term 'labour economist' is heresy. Nevertheless, here's some analytical sense.

His major point is the elephant that's been sitting in the room from the beginning: if labour productivity rises, how much of that productivity increase accrues to workers? Or is it the case, as it has been since at least Eli Whitney, that more productivity leads to labour shedding? After all, that's the express point, now isn't it?

Sam Altman of the ChatGPT maker OpenAI sees A.I. wiping out poverty.

That is just about the stupidest, ahistorical statement that anyone can make. How, exactly, are the poor, uneducated, rednecks living in Appalachian shacks (or worse, your average urban ghetto) going to be transformed to 90210 (RIP, Shannen)? The answer, of course, is not.

Years ago I worked for a regional store-front computer training outfit, which had been a straightforward COBOL school in real brick and mortar facilities with real IBM Big Iron. But then the Damn Gummint dangled moolah to 're-train' welfare folks, mostly mothers, for word processing and Lotus 1-2-3 office jobs. Turned out that almost all of my mothers were there under court order; skip school and ya don't get your kid(s) back. Needless to say, they evinced little enthusiasm for improving their skills. Ultimately, the outfit was found out and shuttered. Little of the moolah went into the classroom, and most went into the owner's pocket. And, also, an MIT grad. He wore his class ring with distinction.

An earlier essay in these musing referenced The Musk Ox's assertion that AI and the like mustneeds develop in concert with a massive re-distribution of income and wealth in order for the machine to keep running. Naturally, The Musk Ox was gleefully dis-employing his own workers rather than moving forward with such re-distribution. But, as Eccles observed 70 years ago: the machine stops working when income and wealth get so lopsided that aggregate demand collapses.

One of Acemoglu's critics puts it this way

"A lot of the benefits of A.I. will come from getting rid of the least productive firms," argues Tyler Cowen, an economist who says the model behind Acemoglu's study is wrong.

Of course, Cowen (a raving Evangelical Radical Right Wingnut of the first water) may be correct, so far as it goes. But if significant numbers of American business is shuttered, how do the dis-employed continue to buy the cornucopia of burgeoning productivity? They don't, stupid. Which is a feature, not a bug. The Evangelical Radical Right Wingnuts seek to return to the thrilling days of yesteryear, when there were a few elite with all of the creature comforts and the serfs making do with subsistence. They just dress it up in fancy verbiage.

But, it remains the case (as of today) as others and I have asserted, AI as currently implemented is just a massive correlation matrix of internet shit. Some call it search on steroids. That, too. `