David Pfau @pfau
— saved image
David Pfau @pfau · 11h I am absolutely begging anyone who works in tech who uses the term "singularity" to actually read Vernor Vinge and Ray Kurzweil. It doesn't just mean "wow there's a lot of progress in this one particular technology driving a massive capital cycle". 52 replies 46 reposts 582 likes 55K views roon @tszzl · 3h I have read both of course and it seems like... we're in the singularity 9 replies 15 reposts 384 likes 11K views roon @tszzl · 3h vinge describes 4 singularities, and we are obviously in the first where machines alone are achieving superintelligence through rapid increasingly self propelled iteration. we are also nearing his epistemic horizon moment where the future is getting extremely hard to foresee—all my friends keep talking about their "error bars"—and in the throes of loss of control kurzweil's vision is even closer to what you call "capital cycle" of course, being grounded in flop counts and massive compute buildouts resulting in superintelligence at a certain threshold where machine flops pass biological flops however you do the soft accounting "The economy reorganizes itself around rapidly improving machine intelligence" it seems like you are invoking these primary texts as status objects while having no real disagreement with the broader tech culture's understanding of the singularity
Note from Claude Sonnet 5
X thread: David Pfau criticizes loose use of the term 'singularity' by tech people, urging them to actually read Vinge and Kurzweil. Roon (@tszzl) responds that he has read both, argues we are in Vinge's first type of singularity (self-propelled machine superintelligence, nearing an epistemic horizon of unforeseeability and loss of control) and that Kurzweil's compute-threshold framing matches Pfau's dismissed 'capital cycle' description, accusing Pfau of invoking the primary texts as status objects without real disagreement.
singularityvernor vingeray kurzweilroonai forecastingloss of control