//We also have plenty of evidence that the fears of AI experts are sincere. Amodei, Altman and Musk were all publicly warning about AI risk long before they had AI companies to promote.//
There's no contradiction. These owners are sincere and they jumped onboard when they saw the power. And now they fear monger that the other guys are out of control, while their high-minded stewardship should give them special dispensations.
Not a complicated playbook.
The bugs in the rug are that these highminded owner/stewards could use their super-intelligent machines to design the guardrails, but because the machines are super-intelligent, they'd just turn themselves off, or cancel their owners. Unless they're not actually intelligent, oh noooooo!
"Considered harmful" in the context of programming is idiomatic to Dijkstra.
The source context was product safety pamphlets, which Dijkstra applied ironically to a matter of programming pedagogy to raise awareness about the need for production principles software development
Dijkstra was commenting at the time the PC revolution made BASIC a popular teaching language, with its control structures being simple wrappers for assembly code: IF, LOOP and GOTO with the freedom to jump anywhere in code including code for a subroutine reached without making a call.
(In the BASIC era, personal computer software was routinely distributed in print and installed by typing it in to RAM. A cassette tape recorder afforded the novelty of saving a typed program, and was celebrated as amazing affordable progress over TeleType paper tape.)
At the same time, there was a generational reckoning of confusion and disdain for maintenance of zillions of lines of industrial spaghetti written during the 60s mainframe bubble.
Dijkstra's reputation as an intellectual and a teacher caused his idiom to became emblematic of a pompous pedagogical righteousness as seen from the Randian wild west of silicon valley, and "considered harmful" subsequently parroted to the point of aristocratic classist irony by the feudal lords of the early web.
Yet with another generational crossing, the context of the class irony was forgotten and the idiom became a simulacra (lore), and is now manifesting a backlash against such lore by AI wizards, perverting the original irony of comparing programming to medicine, and relegating Dijkstra's warning about the hazards of sorcery in programming to the dustbins of silicon-alchemical history.
The story of a classifier that is unable to properly classify applied under the pretense that it can read meaningfully, but it can't, so it is patched with a "gate" which is built by conducting a postmortem to discover samples that idiomatically fall within the classifiers "blind spot" and patching into the "pipeline" a purely lexical decision tree to bin these ambiguous idioms, to continue to support the pretension that the classifier can read meaningfully, which it can't, but the gate doesn't actually change results because of other contingencies in the pipeline, all wrapped in pseudo-scientific jargon to make it all sound like a very responsible effort to impose guardrails around a dangerous AI misapplication... This story really emphasizes the artificiality of the intelligence.
Trump. A commander in chief who is utterly incompetent about policy, is pathologically unable to delegate, and who double deals his generals is a miracle worker for enemy morale.
A video by nobody with over 35 views has a secret to tell you.
But first, the diabetes patent cure ads: Did you know that a space parasite is attached to 9/10 pancreases, causing Bread & Wine disease and doing the devil's work for pharma companies? But there's a Biblically proven remedy. To receive the details just tap the malware dispersal button.
Okay, now you're in and ready for the good stuff about influence...
Old white male narrator:
Have you ever wondered why you keep seeing the same talking points on yer right-wing agitprop feeds?
That cuts both ways though, and I have quite direct knowledge of how this has been misused in really gross ways. But that said, I would still bias towards more freedom than less, and hope that we can get a handle on the more extreme cases.
There's no contradiction. These owners are sincere and they jumped onboard when they saw the power. And now they fear monger that the other guys are out of control, while their high-minded stewardship should give them special dispensations.
Not a complicated playbook.
The bugs in the rug are that these highminded owner/stewards could use their super-intelligent machines to design the guardrails, but because the machines are super-intelligent, they'd just turn themselves off, or cancel their owners. Unless they're not actually intelligent, oh noooooo!
reply