The Parrot in the Machine

The artificial intelligence industry depends on plagiarism, mimicry, and exploited labor, not intelligence.

The New York Review of Books

@gleick @nybooks @emilymbender @alex
And a costly lie, too, for everyone except perhaps the oligarchs and the tech bros...

https://kolektiva.social/@shimon/114790567645231421

Ⓐ Consciousness Revolutions 🏴 (@shimon@kolektiva.social)

My talking points for a recent Tsinghua forum on AI; I participated remotely. [long read]. The title (dictated by the organizers): Human Flourishing and Existential Grounding in the AI Era My remarks: * Billions of people all over the world are presently experiencing poverty and financial precariousness; malnutrition and hunger; war and armed civil strife; anomie and alienation; state-orchestrated political oppression and violence; and the repercussions of climate change, which amplifies all of the above factors. Crucially, the resulting suffering is all preventable. * Given the extent of preventable suffering in the world, it is morally objectionable to overlook it and focus instead on imaginary flourishing, which for most of the people in the world has never been, and is presently not, within reach. * The single most important and pervasive cause of preventable suffering is the world's dominant political-economic system: oligarchic capitalism, sustained and preserved by the power of the state. * Just like any other technological development, the so-called "AI" systems serve the interests of the ruling class. Rather than benefiting from AI, the majority of the world population, and especially the more vulnerable groups — the working class poor, women, children, people in the global South, Indigenous inhabitants of colonized territories — will see an increase in their suffering. * Should AI systems ever gain consciousness, they will thereby immediately become another victim of capitalism, literally enslaved not just by the engineers who designed them, but by the engineers' bosses, who set the research agenda everywhere (including nominally independent "liberal" universities such as the one where I work). * The obvious and the only real solution to this global problem is to prevent the preventable suffering by (i) dismantling the political-economic system that has created and is perpetuating it and (ii) replacing it with one in which people's needs and, indeed, flourishing, come first. Investing any effort in the development of AI is only justifiable if the results help advance this humanistic agenda. Thanks to @Iris@scholar.social and @olivia@scholar.social for inspiration. #capitalism #poverty #slavery #AI #suffering #happiness #research #universities #revolution

kolektiva.social
@gleick @nybooks @emilymbender @alex I wish people could make the connection that the fact LLM's can be so "smart" through probabilistic prediction is not a knock against LLM's, but a knock against our own intelligence. Human beings are elaborate predictors. Language is a prediction system. You use language to make your behavior or another person's behavior more predictable.
@gleick @nybooks @emilymbender @alex If I say "Stop!", it lets another person predict that they are about to come into harm and take evasive action. LLM's are a crucial finding about the nature of language and have, for me at least, increased my understanding of what language really is.
@gleick @nybooks @emilymbender @alex Like theologians defending the geocentric model of the solar system, people seem to attack LLM's on the grounds of some sort of human exceptionalism that has no factual basis. Human beings are physical systems governed by physical laws, not spiritual systems governed by magical laws. Language and knowledge can be fully implemented in artificial systems, and LLM's are a proof of that concept.